NIPS 2012 papers

(in nicer format than this)
maintained by @karpathy
source code on github
Below every paper are TOP 100 most-occuring words in that paper and their color is based on LDA topic model with k = 7.
(It looks like 0 = theory, 1 = reinforcement learning, 2 = graphical models, 3 = deep learning/vision, 4 = optimization, 5 = neuroscience, 6 = embeddings etc.)
Toggle LDA topics to sort by: TOPIC0 TOPIC1 TOPIC2 TOPIC3 TOPIC4 TOPIC5 TOPIC6
Random Utility Theory for Social Choice
Hossein Azari, David Parks, Lirong Xia


[order, ranking, algorithm, set, theorem, bounded, case, probability, partial, true, lemma, consider, path, maximize, theory, sense, provide, chosen, subset, adopt] [approach, utility, choice, function, preference, focus, expectation] [likelihood, normal, model, number, social, distribution, inference, maximum, alternative, monte, compute, election, carlo, parameter, predictive, gibbs, exj, lirong, public, bayesian, exponential, gumbel, rum, full, latent] [global, figure, ground, location, truth, corresponds, evaluate, apply, randomly] [random, condition, rank, problem, suppose, method, correlation, concave, estimated, theoretical, kendall] [mle] [data, performance, better, dataset, variance, synthetic, linear, learning]
Adaptive Stratified Sampling for Monte-Carlo integration of Differentiable functions
Alexandra Carpentier, Remi Munos


[algorithm, stratum, lemma, bound, oracle, order, asymptotic, measure, consider, uniform, lower, provide, crude, allocates, differentiable, assume, prove, proof, uniformly, write, sequence, quantity, budget, smaller, access, diameter, call, admits, fact, result, asymptotically, theorem, stratify, paper, main, rounding, probability, negligible, aim, carpentier, discrepancy, allocating, allocate] [function, optimal, domain, strategy, intuition, taylor, good, adapt, provided] [sampling, number, sample, partition, monte, carlo, supplementary, expansion, will, integral] [larger, build] [estimate, assumption, random, quasi, constant, rate, low, problem] [adaptive, material, integration] [squared, error, point, variance]
Causal discovery with scale-mixture model for spatiotemporal variance dependencies
Zhitang CHEN, Kun ZHANG, Laiwan CHAN


[lemma, algorithm, assume, note, true, series, energy] [assumed, observation, conference, based] [model, sample, observed, log, ordering, structure, discovery, latent, gaussian, factor, variable, component, generated, confounders, place] [temporal, figure, structural, randomly, discovered, level, second] [causal, contemporaneous, independent, lingam, spatiotemporal, stock, matrix, estimate, assumption, estimation, estimated, method, propose, size, causality, investigate, journal, acyclic, autoregressive, dft, problem, zhang, rxy, multivariate, ftse] [time, temporally, mechanism, neural, signal, svar] [variance, linear, data, proposed, selected, real, vector, learning, table, machine, synthetic, large]
A Bayesian Approach for Policy Learning from Trajectory Preference Queries
Aaron Wilson, Alan Fern, Prasad Tadepalli


[query, set, active, measure, algorithm, consider, change, length, learner, energy, disagreement, comparison, example] [policy, expert, state, trajectory, function, agent, tpq, bicycle, initial, tpqs, expected, preference, approach, action, based, evaluation, strategy, acrobot, domain, car, generating, control, goal, cart, mountain, reward, pole, knowledge, start, reinforcement, describe, starting, current, benchmark] [distribution, posterior, model, generated, generate, pair, latent, sampling, prior, bayesian, hmc, inference, sample, number, variational, will] [target, work, simple, produce, learn, feature, map] [random, gradient, problem, estimate, step] [response, desired, selective, velocity] [learning, selection, select, data, space, performance, machine, vector, distance, dissimilarity]
Training sparse natural image models with a fast Gibbs sampler of an extended state space
Lucas Theis, Jascha Sohl-Dickstein, Matthias Bethge


[algorithm, statistical, probability, set, best, case, change, variant] [optimal, transition, average, additional, international, based, standard] [overcomplete, model, sampling, gibbs, distribution, posterior, sampler, markov, gaussian, blocked, hmc, will, chain, nullspace, inference, laplace, mala, olshausen, autocorrelation, hidden, likelihood, number, computational, monte, sample, latent, marginals, overcompleteness, chen, carlo, prior, marginal, gaussians, log, pot] [source, image, figure, training, gsm, single, contrast, representation, local] [sparse, convergence, journal, estimate, toy, constrained, dkl] [basis, natural, complete, persistent, neural, study, time, making] [linear, learning, data, advantage, machine, better, performance, point]
Neurally Plausible Reinforcement Learning of Working Memory Tasks
Jaldert Rombouts, Sander Bohte, Pieter Roelfsema


[set, probability, highest] [reward, decision, action, reinforcement, correct, required] [model, probabilistic, assigned] [layer, trained, learn, target, training, learns] [convergence, stochastic] [memory, synaptic, association, working, activity, feedback, sensory, cue, network, unit, augment, regular, signal, cortex, trial, neural, input, neuromodulatory, feedforward, activation, biologically, red, plausible, relevant, mark, evidence, motor, time, represent, green, mechanism, computed, agjk, dopamine, persistent, vlm, winning, agij, responsible, left, integrate, shaping, tuned, plasticity, nature, explains, aglm, attentional, difference, presented] [learning, task, scheme, mapping, error, binary]
Deep Representations and Codes for Image Auto-Annotation
Ryan Kiros, Csaba Szepesvari


[set, denote, algorithm, best, follow, introduce, length] [standard, approach, tile, average] [number, model, hierarchical, sample, collection, base] [image, training, module, annotation, feature, second, perform, deep, representation, layer, learned, cross, figure, multiple, tagprop, single, spatial, unsupervised, matching, pooling, top, color, larger, coates, pooled, apply, performed, retrieval, pixel, extract, stride] [dictionary, size, extraction, optimization, sparse, successful] [natural, coding, input, receptive, tag] [learning, distance, performance, dataset, test, existing, large, small, linear, selection, metric, binary, table, hamming, code, vector, nearest, construct]
Fiedler Random Fields: A Large-Scale Spectral Approach to Statistical Network Modeling
Antonino Freno, Mikaela Keller, Marc Tommasi


[set, statistical, order, probability, respect, suitable, result, paper, denote, prove, divergence] [approach, parametric, function, neighboring] [fiedler, xuv, model, guv, frfs, structure, frf, distribution, concerning, delta, modeling, number, degree, pair, observed, node, conditional, form, vuv, mlp, link, involved, sample, sampling, variable, markov, allows, euv, subgraphs, xwz, undirected, prior, aimed, unwarranted, nonparametric, hepth] [accuracy, training, contrastive, target, work, second] [graph, random, dependence, size, edge, eigenvalue, subgraph, spectral, estimate, assumption, arxiv, estimation, proposition, estimating, objective] [network, prediction, potential, basic, connected] [learning, laplacian, statistic, data, smallest]
A Unifying Perspective of Parametric Policy Search Methods for Markov Decision Processes
Thomas Furmston, David Barber


[considered, term, algorithm, notation, consider, note, property, denote, example] [approximate, policy, ascent, expectation, state, reward, maximisation, function, search, parametric, decision, expected, total, mdp, standard, reinforcement, planning, trajectory, current, action, control, transition, wmean, provided, choice, agent] [parameter, form, will, markov, log, update, supplementary, full, distribution, inference] [experiment, training, performed, second, local] [newton, gradient, method, steepest, objective, matrix, diagonal, analysis, quadratic, optimisation, desirable, plot, hessian, application, optimum, convergence, derivative] [natural, system, direction, fisher, dynamic, plotted, negative, difference] [space, linear, learning, machine, performance, error, large]
Tight Bounds on Profile Redundancy and Distinguishability
Jayadev Acharya, Hirakendu Das, Alon Orlitsky


[redundancy, probability, lemma, def, bound, sequence, length, universal, upper, multiset, alphabet, minimax, distinguishable, consider, induced, class, set, testing, lower, uniform, minimum, element, result, easy, theorem, multiplicity, denote, regret, proof, closeness, divergence, assign, diminishing, hypothesis, online, uniformly, consists] [expected, unknown, describe, choice] [distribution, log, number, collection, generated, poisson, partition, discrete, sampling, will] [ieee, pattern, sum, learned] [entropy, random, size, compression, rate, needed, essentially, estimation, compressed] [encode, determined, increase, coding, symbol, power] [large, learning, code, largest, binary, machine, improving, data, well]
Active Learning of Model Evidence Using Bayesian Quadrature
Michael Osborne, David Duvenaud, Roman Garnett, Carl Rasmussen, Stephen Roberts, Zoubin Ghahramani


[active, assume, true, note, statistical, choose, term, set, algorithm] [function, expected, approach, uncertainty, approximate, standard, decision] [log, bayesian, posterior, likelihood, hyperparameters, model, prior, quadrature, gaussian, sampling, monte, bbq, sample, integral, carlo, number, will, clog, marginal, smc, inference, university, mlog, chain, exp, distribution, marginalisation, marginalising, integrand, evaluating, linearisation, compute, computing, permitted, mlii, bmc, modelling, hyperparameter, ale] [figure, evaluate, scale, second, previous, work, location, truth, single] [method, estimate, numerical, convergence, estimated, high, covariance, novel, independent] [evidence, integration, input, tested, computed, negative] [variance, small, learning, real, synthetic, select, data, better, table]
Relax and Randomize : From Value to Algorithms
Sasha Rakhlin, Ohad Shamir, Karthik Sridharan


[relt, online, rademacher, inf, algorithm, regret, complexity, minimax, bound, loss, lemma, consider, follow, upper, learner, transductive, leader, perturbed, version, set, supremum, classical, bounded, case, class, sequence, provide, example, respect, uniform, simultaneously, proof, regt, recipe, playout] [admissible, sequential, round, strategy, optimal, idea, game, based, expected, function, expectation] [distribution, conditional, draw, exp, sampling, exponential, log, will, develop, development] [view] [relaxation, random, convex, matrix, assumption, problem, norm, method, proposition, argmin, optimization, involves, completion, constant, suppose, descent, entry] [unit, prediction, study, static, nature, equation, potential] [learning, randomized, trace, ball, space, linear, admissibility, binary]
Bayesian nonparametric models for ranked data
Francois Caron, Yee Whye Teh


[probability, partial, list, measure, note, ranking, consider, derived, set, extension, statistical, theorem, case, write, paper, derive, easily, branching, series, algorithm] [choice, approach, continuous, simply, total, desirability, atomic] [model, gamma, nonparametric, posterior, bayesian, gibbs, process, law, will, update, parameter, conditional, atom, distribution, prior, observed, sampler, mass, inference, exp, wtk, number, joint, described, latent, ctk, arrive, york, modeling, sampling, mutually, sample, completely, discrete, develop, form, associated] [figure, simple, appear, previous] [random, journal, dependence, normalized, positive, independent, distributed, corresponding, suppose, stochastic] [time, auxiliary, ith, dynamic, limit] [item, rating]
Privacy Aware Learning
John Duchi, Michael Jordan, Martin Wainwright


[loss, statistical, theorem, lower, assume, upper, set, risk, minimax, theory, case, guarantee, measure, inf, version, denote, datum, access, extreme, sharp, main, bound, corollary, symposium, probability, class, radius, perturbed, bounded, consider, supported, survey] [optimal, function, saddle, approach] [distribution, conditional, log, sample, collection, parameter, allows] [local, second, view, level, characterize, apply] [stochastic, mutual, convergence, random, estimation, convex, method, gradient, proposition, subgradient, procedure, descent, journal, supp, rate, entropy, norm] [natural, study, amount, channel, regular] [privacy, data, learning, private, point, compact, differential, error, differentially, approximation]
Feature-aware Label Space Dimension Reduction for Multi-label Classification
Yao-Nan Chen, Hsuan-Tien Lin


[label, loss, algorithm, bound, minimizes, set, called, minimizing, equivalent, consider, design] [approach, optimal, conference] [number, conditional, parameter, component, enron] [feature, figure, transformation, canonical, training, scene, medical, unsupervised, three] [matrix, problem, correlation, min, analysis, singular, optimization] [prediction, encoding, principal, effective, experimental, coupled, decoding] [cplst, plst, error, regression, lsdr, space, linear, kernel, zvt, dimension, hamming, learning, occa, cca, reduction, better, ridge, vector, table, test, supervised, fsdr, machine, performance, dataset, regularization, zwz, vvt, pbr, considers, binary, xwx, code, orthogonal, national, taiwan, original, row, worse, small, yeast, data]
Q-MKL: Matrix-induced Regularization in Multi-Kernel Learning with Applications to Neuroimaging
Chris Hinrichs, Vikas Singh, Jiming Peng, Sterling Johnson


[rademacher, complexity, set, note, margin, bound, result, derived, theorem, arbitrary, comparison, mixing, order, setting, algorithm] [standard, domain, state, function, knowledge] [base, model, will, disease, structure, cluster, form] [feature, multiple, performed, imaging, image, directly, combined, sum, trained, accuracy, spatial, module, exploit] [problem, entropy, norm, optimization, quadratic, penalty, convex, method, min, dual, solve, program, density, objective, solution, support] [prediction, effectively, brain, optimized] [kernel, mkl, learning, virtual, renyi, svm, combination, vector, relative, generalizes, space, neuroimaging, error, large, linear, data, proposed, regularizer, constraint, regularization, relates, tbm, art, laplacian]
Multiresolution Gaussian Processes
Emily Fox, David Dunson


[cut, set, assume, consider, statistical, smoothness, series, empirical] [function, based, key, standard, approach, cost] [partition, mgp, tree, hierarchical, likelihood, gps, process, parent, gaussian, predictive, heldout, abrupt, marginal, meg, proposal, posterior, distribution, generated, structure, model, log, sharing, prior, allows, inference, conditioned, word, collection, changepoints, bayesian, hgp, sample, conjugacy, modeling, mcmc] [global, level, figure, local, capture, compare, temporal, single] [covariance, normalized, correlation, analysis, smooth, matrix, formulation] [time, visual, input, trial, additive, neural] [data, shared, multiresolution, functional, point, binary, common, proposed]
A P300 BCI for the Masses: Prior Information Enables Instant Unsupervised Spelling
Pieter-Jan Kindermans, Hannes Verschore, David Verstraeten, Benjamin Schrauwen


[online, set, probability, algorithm, best, consider] [approach, correct, knowledge, based, start, average] [language, model, will, prior, number, forward, bayesian, full, update, backward, initialization, distribution] [character, spelling, unsupervised, subject, eeg, tri, training, speller, uni, trigram, trained, figure, train, work, single, spell, wold, second, supervisedly, ieee, spelled, accuracy, computer, incorporating, period, tatri, previous, akimpech, randomly, compared, incorporation] [method, application, graphical, pass] [transfer, time, desired, weight, bci, brain, adaptive, neural, interface, adaptation, limited, computed, basic, processing] [learning, dataset, test, proposed, supervised, performance, text, data, table, vector, linear, letter]
Learning about Canonical Views from Internet Image Collections
Elad Mezuman, Yair Weiss


[hypothesis, called, algorithm, provide, set, true, returned, decide, change, example] [search, conference, international, simply, preference] [number, will, distribution] [view, canonical, image, frequent, object, gist, human, internet, computer, mode, palmer, category, psychophysical, figure, automatic, vision, parzen, second, recognition, descriptor, imagenet, ieee, viewpoint, corresponds, manually, formal, google, bounding, work, cognitive, experiment, retrieved, maximal, horse, center, cropped, reveal, shape, downloaded, pattern, iconic, proxy, learn, notre] [method, density, strong, estimation] [preferred, phase, presented, visual, frequency, basic, output, perception] [similarity, small, clustering, large, verify, space, manual, dataset, data]
MCMC for continuous-time discrete-state systems
Vinayak Rao, Yee Whye Teh


[interval, set, algorithm, probability, length, sequence, smaller, note, observe] [state, trajectory, function, observation, approach, cost, continuous, conference, temperature] [sampler, smjp, mcmc, particle, hazard, markov, poisson, process, thinning, thinned, sampling, event, jump, ass, tend, discretization, sample, inference, number, distribution, uniformization, parameter, dominating, posterior, chain, leaving, pmcmc, model, monte, bvi, likelihood, will, computational, weibull, resample] [second, figure, simple, three, scale, applied] [rate, random, distributed, constant, independent, step, general, truncated, proposition, stochastic] [time, system, effective, unit, waiting, represent, population, long] [candidate, large, equal]
Learning High-Density Regions for a Generalized Kolmogorov-Smirnov Test in High-Dimensional Data
Assaf Glazer, Michael Lindenbaoum, Shaul Markovitch


[set, ocsvm, ocnm, theorem, change, respect, statistical, bep, note, polonik, version, israel, introduce, annals, measure, reference, distributional, fraction, algorithm, hold, minimum, probability, best, asymptotically, quantile, assume] [approach, decision, based, function, expected, evaluated, average] [hierarchical, distribution, nonparametric, number, associated, sample] [training, detection, figure, precision, level, region, previous, recall, second] [density, method, estimation, estimator, estimated, estimate, generalized, practical, solution, multivariate, estimating, symmetric, sparsity] [input, difference, time, represent] [test, data, learning, kernel, space, distance, averaged, equal, neighbor]
Affine Independent Variational Inference
Edward Challis, David Barber


[bound, class, logistic, set, respect, lower, consider, example, true, partial, complexity, term, divergence, probability] [approximate, standard, optimal, expectation, sampled, provided, function] [gaussian, base, log, inference, variational, bayesian, distribution, likelihood, number, site, prior, whilst, ntrn, qud, model, laplace, bai, marginal, posterior, supplementary, computational, integral, form, latent, skewness] [bounding, training, figure, transform, target, local, evaluate] [density, multivariate, method, dimensional, sparse, robust, entropy, stable, independent, jointly, procedure] [lattice, neural, processing, presented, computation, equation, difference, accurately] [regression, approximation, fourier, linear, approximating, data, test, machine, special, learning, kernel]
Algorithms for Learning Markov Field Policies
Abdeslam Boularias, Oliver Kroemer, Jan Peters


[set, algorithm, approximated, empirical, setting, probability] [policy, function, reward, optimal, bellman, state, approach, expected, reinforcement, apprenticeship, conference, search, action, friction, amn, deterministic, domain, knowledge, maxent, pendulum, provided, grasp, based, decision, international, cart, temperature] [distribution, markov, will, structure, indicates, sampling, inference, process, likelihood, exp] [feature, learned, object, track, handle, figure, structured] [matrix, random, problem, graph, high, entropy, inverse, corresponding, low, proposition, estimated, solution, successful, adjacency, solving, gradient, lagrangian] [equation, hand, surface, robot, basis, grasping, framework] [learning, error, kernel, point, machine, similarity, linear, proposed, space]
The representer theorem for Hilbert spaces: a necessary and sufficient condition
Francesco Dinuzzo, Bernhard Schölkopf


[theorem, lower, term, admits, set, denote, bounded, example, class, sequence, consider, minimized, characterization, proof, result, respect, observe, property, easy, mathematical, differentiable, assume, subset, probability, exists, loss, setting, prove] [choice, function, evaluation, focus] [form, component, number, allowing] [angle, second, aligned, convolution] [family, condition, type, minimizer, norm, problem, general, support, min, closed, subspace, analysis] [broad, study] [regularization, representer, functionals, linear, space, functional, hilbert, kernel, vector, nondecreasing, admit, semicontinuous, data, reproducing, radial, learning, existence, uniquely, ball, machine, point, orthogonal, large, rkhs]
Coding efficiency and detectability of rate fluctuations with non-Poisson neuronal firing
Shinsuke Koyama


[divergence, empirical, lower, bayes, bound, lim, probability, result, derived, assume, statistical, introduce, approximated, note, order, depends, algorithm, interval, applicable, share] [function, good] [log, process, gamma, distribution, latent, degree, gaussian, posterior, parameter, formula, prior, model, generated, exp, examine, poisson] [three, figure, shape, temporal, scale, variation, larger, train, map, feature] [rate, constant, estimation, method, density, journal, numerical, family, theoretical, analysis, inverse, sin, estimate, achieves, stochastic] [spike, isi, lognormal, renewal, neural, time, encoding, cortical, represent, represents, spiking, phase, solid, decoding, determined, neuronal, decoder, ith, coding] [large, space, small]
Why MCA? Nonlinear sparse coding with spike-and-slab prior for neurally plausible image encoding
Jacquelyn Shelton, Philip Sterne, Joerg Bornschein, Abdul-Saboor Sheikh, Joerg Luecke


[algorithm, set, active, main] [standard, optimal, average, function, observation, approach, max, transition] [model, prior, posterior, latent, distribution, generative, gibbs, inference, observed, sampling, gaussian, variational, inferred, computational, log, vivo, generated, component, bayesian, form, laplace, hidden, infer, likelihood, number] [image, simple, multimodal, single, combined, figure, sum, learned, work, maximal, training] [sparse, high, sparsity, exact, journal, dictionary, method, optimization] [coding, nonlinear, neural, visual, globular, receptive, natural, bimodal, latents, preselection, wdh, encoding, basis, superposition, study, gabor, noise, processing, left, emsteps] [data, learning, linear, combination, point, machine]
Learning optimal spike-based representations
ralph Bourdoukan, David Barrett, Christian Machens, Sophie Deneve


[loss, derive, assume, case, term, set] [optimal, function, start, converges] [will, identical, variable] [single, representation, train, learn, simple, produce, target, bottom, work, previous] [rate, balance] [neuron, spike, network, rule, spiking, membrane, input, potential, synaptic, neural, equation, connectivity, balanced, plasticity, hebbian, represent, inhibitory, population, signal, time, excitatory, homogeneous, voltage, output, cortical, reset, neurosci, difference, connected, connection, presynaptic, dynamical, interpreted, weakly, strength, autapse, regime, excitation, stdp, interpret, producing, ith, weight, threshold, optimally, middle, recurrent, inhibition, asynchronous, deriving, correlated, postsynaptic] [learning, error, linear, large, small, point]
"On the (Non-)existence of Convex, Calibrated Surrogate Losses for Ranking"
Clément Calauzènes, Nicolas Usunier, Patrick Gallinari


[surrogate, ranking, loss, calibrated, risk, set, theorem, calibration, respect, err, prove, lemma, pairwise, consider, result, proof, characterization, implies, assume, sort, induced, exists, algorithm, order, main, subset, implication, empirical, dag, studied, measure, induces, consequence, sigir, instance, proved, relevance] [evaluation, optimal, function, search, based, optimizing, average, approach, expected] [distribution] [scoring, three, directly, score] [convex, arg, rank, assumption, satisfy, denoted, general, analysis, minimization, low, min, optimization, technical] [feedback, framework, study, preferred, predicted, lead] [learning, metric, item, vector, inner, quality, better, point, table, sorting, permutation]
Dynamic Pruning of Factor Graphs for Maximum Marginal Prediction
Christoph Lampert


[probability, set, case, example, main, note, runtime, statistical, pairwise, label, stopping, situation, ordinary, introduce, energy] [decision, function, pruning, approximate, correct, incomplete, based, uncertain] [sampling, marginal, sample, factor, number, gibbs, inference, variable, mmp, model, distribution, will, parameter, maximum, bayesian, joint, marginals, sampler, beta, conditional, computational, form, monte, chain, markov] [image, figure, training, accuracy, inpainting, learned, trained, pixel] [graphical, estimate, size, method, close, random, independent, estimated, graph, exact, propose] [adaptive, prediction, effective, early, time, difference, making] [binary, large, dataset, individual, learning, proposed, regularization, small]
A Generative Model for Parts-based Object Segmentation
S. M. Ali Eslami, Chris Williams


[john, algorithm, chosen, probability, report, set, upper, lower] [average, simply] [model, inference, generative, probabilistic, hidden, joint, number, likelihood, chain, multinomial, fully, mixture, distribution, highly, generate, latent, comparable, split, performing, conditional] [shape, appearance, object, msbm, training, segmentation, boltzmann, image, ieee, pixel, sli, figure, sbm, fowlkes, nicolas, foreground, second, local, computer, body, trained, train, scoring, accuracy, thomas, layer, background, belonging, learned, work, learn, ism, vittorio, top, ethz] [random, strong] [account, variability, differ] [dataset, learning, test, data, machine, performance, datasets, labeled, quality, seed, table]
Local Supervised Learning through Space Partitioning
Joseph Wang, Venkatesh Saligrama


[empirical, reject, loss, multiclass, perceptron, logistic, surrogate, example, label, piecewise, consider, complexity, class, lemma, upper, minimize, comparison, algorithm, minimizing, approximated] [decision, approach, based, function, conference, optimizing, observation, cost, international] [number, lda, computational, associated, attempt] [local, training, region, global, trained, complex, feature, simple, discriminant, single, train, indicator, learn, randomly, figure, contrast] [problem, method, analysis, coordinate, family, descent, step, composite] [voted, linearly, stage, output] [linear, learning, partitioning, space, error, data, binary, test, supervised, locally, machine, scheme, proposed, boosting, regression, large, well, datasets, small, performance, optimize, adaboost, choosing, robert]
Learning with Target Prior
Zuoguan Wang, Siwei Lyu, Gerwin Schalk, Qiang Ji


[sequence, set, algorithm, length, regressor, online, loss] [parametric, function, based, knowledge, expectation, approach] [prior, model, parameter, joint, latent, bayesian, form, variable, probabilistic, log, distribution] [target, ltp, subject, training, unsupervised, pose, learned, ecog, learn, sgplvm, corresponded, image, train, temporal, directly, trained, uncorresponded, work, figure, structured, paired, second, body, electrocorticographic] [corresponding, estimation, batch, random, general, conventional, estimate] [decoding, bci, input, brain, signal, movement, time, output, tested, collected, prediction, predict, framework] [learning, data, linear, regression, well, task, proposed, supervised, ridge, generalize, space, table, demonstrate, mapping]
The variational hierarchical EM algorithm for clustering hidden Markov models.
Emanuele Coviello, Antoni Chan, Gert Lanckriet


[algorithm, lower, bound, sequence, probability, derive, set, respect, smaller, paper, series, divergence] [state, observation, expected, max, expectation, based] [hmm, variational, hmms, hidden, model, markov, mixture, log, cluster, hierarchical, vhem, music, hem, distribution, likelihood, base, component, gaussian, will, generated, emission, number, variable, collection, derivation, generative] [level, gmm, annotation, learned, learn, semantic, second, ieee, directly, automatic, capture, speech, entire] [novel, estimation, spectral, reduced, matrix, estimated, support, intermediate] [motion, time, input, group, represents] [clustering, learning, data, virtual, table, machine, large, better, zij, original]
Optimal Regularized Dual Averaging Methods for Stochastic Optimization
Xi Chen, Qihang Lin, Javier Pena


[algorithm, set, online, empirical, loss, assume, extension, corollary, proof, term, probability, note, theorem, convexity, bregman] [optimal, deterministic, average, based, function, conference, expected, current] [parameter, log, number, allows, university] [figure, directly, previous, achieved, improves] [convex, stochastic, rate, orda, convergence, averaging, dual, objective, assumption, gradient, subgradient, optimization, proximal, achieves, regularized, solution, rda, method, achieve, sage, sparsity, strong, solving, smooth, problem, technical, recovery, propose, iteration, avg, fobos, prox, journal, step, high, sparser, accelerated, subgradients, constant, arg] [presented, basic] [learning, table, mapping, machine, approximation, better, proposed, vector, combination]
Parametric Local Metric Learning for Nearest Neighbor Classification
Jun Wang, Alexandros Kalousis, Adam Woznica


[instance, set, algorithm, complexity, term, margin, class, pairwise, lemma, main, bound] [function, cost, optimal] [number, will, model, vary, parameter, predictive] [local, learn, global, learned, learns, single, compare, accuracy, automatic, improves, figure, weighting] [matrix, problem, optimization, method, smooth, dual, objective, norm, gradient, lipschitz] [basis, weight] [metric, learning, plml, distance, linear, performance, anchor, nearest, large, ijk, datasets, kernel, cblml, data, squared, triplet, error, svm, better, wibk, neighbor, glml, wibl, manifold, psd, constructed, lmnn, regularization, mahalanobis, space, isolet, zbl, constraint, kbl, boostmetric, cijk, well, appropriate, row, sml, select, worse, respective, dataset]
Pointwise Tracking the Optimal Regression Function
Yair Wiener, Ran El Yaniv


[lemma, regressor, hypothesis, probability, loss, bound, risk, erm, set, theorem, pointwise, active, coverage, case, reject, regressors, assume, proof, paper, class, result, empirical, including, statistical, substantial, true, algorithm, version, increasing, hindsight, agnostic, rejection, supremum, disagreement] [strategy, optimal, function, standard, optimality, average, unknown, required, based] [sample, will, full] [training, applying, entire, learn, tracking, ieee, feature, region, baseline] [convex, method, constrained, general, monotonically, size, numerical, discussed, matrix, application] [selective, prediction, implementation, neural, input, multiplicative, difference, presented] [learning, regression, linear, test, error, squared, point, proposed, datasets, technique, vector, machine, space, distance, performance, common, row]
Monte Carlo Methods for Maximum Margin Supervised Topic Models
Qixia Jiang, Jun Zhu, Maosong Sun, Eric Xing


[set, term, equivalent, denote, probability, margin, derive] [average, good, function, based, expectation, strategy, goal] [topic, posterior, medlda, distribution, gmedlda, imedlda, model, lda, sample, latent, variational, gibbs, inference, sampling, sampler, monte, carlo, number, bayesian, fmedlda, collapsed, draw, dirichlet, document, markov, likelihood, prior, zdn, glda, will, chain, discovering, wdn, maximum, flda] [training, accuracy, directly, discriminative, larger, map] [estimate, problem, entropy, solve, size, solution, optimum, dual, sparsity, regularized, stable, method, iteration, min] [prediction, interpretation, natural, rule, desired] [data, supervised, performance, learning, large, space, sensitivity, svm, vector, small, test, weaker, criterion]
A Convex Formulation for Learning Scale-Free Networks via Submodular Relaxation
Aaron Defazio, Tiberio Caetano


[submodular, set, algorithm, consider, case, comparison, sort, order, denote, class] [function, approach, optimizing, standard, good, evaluated] [degree, prior, model, structure, node, log, will, distribution, exponential, number, gaussian, form, tractable, parameter] [scale, applied, figure, accuracy, structured, performed] [proximal, method, dual, convex, problem, decomposition, graph, covariance, sparse, subgradient, relaxation, optimization, reweighted, reconstruction, random, graphical, edge, sparsity, matrix, admm, gene, iteration, erg, diagonal, liu, estimation, convergence, mnp, proposition, alternating, ihler, francis, free, solving, objective, subproblem, envelope, eigenvalue, min, arg, regularized, minimization] [network, natural, weight, absolute, practice, connected] [operator, selection, learning, small, machine]
Probabilistic n-Choose-k Models for Classification and Ranking
Kevin Swersky, Danny Tarlow, Ryan Adams, Rich Zemel, Brendan Frey


[ranking, label, logistic, case, set, maximize, algorithm, probability, note, class, main, relevance, consider, inductive, example, chosen] [expected, function, optimal, approach, evaluation, standard, programming, international] [model, prior, number, distribution, inference, log, count, sample, likelihood, will, structure, probabilistic, draw, modeling, assigned, joint, alternative, conditional, compute, ordering, monotonic, maximum, university, softmax] [top, training, work, multiple, train, single, learned, trained, figure, simple, image, clear, include, sum, restricted] [objective, proposition, exact] [gain, dynamic, output, issue] [learning, binary, ordinal, test, regression, machine, data, vector, dataset, regularization, performance]
How Prior Probability Influences Decision Making: A Unifying Probabilistic Model
Yanping Huang, Abram Friesen, Timothy Hanks, Mike Shadlen, Rajesh Rao


[probability, piecewise, order, case] [decision, belief, optimal, reward, function, state, policy, action, pomdp, eye, knowledge, expected, choice, correct, observation, receives, reinforcement, chooses, based, future, observable] [model, prior, distribution, hidden, normal, sampling, bayesian, university, compute, variable, posterior, markov, inference, emission, diffusion, joint] [human, figure, accuracy, target] [random, coherence] [time, animal, making, reaction, experimental, direction, discrimination, neural, motion, carpenter, psychometric, framework, explain, perceptual, movement, rnrs, saccadic, offset, boundary, dynamic, tnd, represents, determined, reciprocal, additive, neutral, maximization, intercept, trial, brain, sensory, latency, rightward, solid, predicts] [data, task, learning, linear, well, biased]
Fused sparsity and robust estimation for linear models with unknown variance
Arnak Dalalyan, Yin Chen


[case, risk, theorem, setting, probability, respect, provide, fundamental, order, chosen, change, remark, proof, inequality, note, property] [standard, unknown, cone, context, evaluation, interested, choice] [maximum, likelihood, number, will, sample, pair, gaussian] [scale, level, subject, computer, restricted, accuracy] [problem, matrix, estimation, robust, fused, lasso, sparsity, estimator, theoretical, condition, dantzig, convex, solution, penalized, selector, sfds, scaled, min, optimization, procedure, estimating, corresponding, diagonal, ave, sparse, minimization, satisfy, xmj, emmanuel, eigenvalue, solved, penalty] [noise, tuning, std, computed] [vector, linear, large, regression, small, table, synthetic, special, quality, learning]
A Linear Time Active Learning Algorithm for Link Classification
Nicolò Cesa-Bianchi, Claudio Gentile, Fabio Vitale, Giovanni Zappella


[set, asymexp, query, spanning, signed, treecutter, active, algorithm, label, theorem, order, path, patht, querying, fraction, queried, learner, arbitrary, smaller, consistent, visit, labeling, fact, slashdot, bound, starmaker, epinions, length, ytrain, integer, setting, bounded, case, subtrees] [optimality, based, conference, average] [tree, number, factor, social, link, log, node, model, undirected, assignment] [training, figure, remaining, larger, randomly, simple, work, created] [graph, edge, size, constant, analysis, spectral, positive, theoretical, achieves, running, stochastic] [time, prediction, sign, connected, predicted, bias, input, negative, network] [learning, test, labeled, predicting, linear]
Complex Inference in Neural Circuits with Probabilistic Population Codes and Topic Models
Jeff Beck, Katherine Heller, Alexandre Pouget


[algorithm, probability, note, online, fact, equivalent] [function, approximate, optimal, simply, standard, approach] [probabilistic, distribution, latent, inference, posterior, model, topic, exp, vbem, log, lda, bayesian, variational, generative, document, form, word, intensity, normalization, number, likelihood, will, dirichlet, targeted, infer, assignment, update, njk] [pattern, complex, build, layer, figure, simple, scene] [problem, general, estimate, matrix] [neural, network, spike, activity, population, implementation, input, divisive, ppc, natural, represents, olfactory, neuron, dynamic, demixing, output, odor, visual, framework, nature, biologically, capable, rule, time, system, implementing, faced, relevant, plausible, ddm] [linear, learning, task, approximation, code]
Truncation-free Online Variational Inference for Bayesian Nonparametric Models
Chong Wang, David Blei


[algorithm, set, complexity, lower, bound, empirical, comparison, consider, case] [approach, international, conference, approximate, intelligence, grows] [variational, inference, mixture, distribution, topic, number, dirichlet, collapsed, likelihood, model, bnp, york, batchsize, bayesian, process, hidden, latent, gibbs, document, component, hdp, posterior, update, sampling, log, hierarchical, truncation, word, parameter, nonparametric, exp, prior, sample, inferred, predictive, massive, vocabulary] [local, global, figure, work] [stochastic, method, batch, journal, size, suppose, unbounded, rate, family, random] [nature, time, neural, processing, natural, difference] [data, learning, machine, locally, large, better, small]
Exact and Stable Recovery of Sequences of Signals with Sparse Increments via Differential _1-Minimization
Demba Ba, Behtash Babadi, Patrick Purdon, Emery Brown


[sequence, theorem, proof, fact, consider, case, set, bound, denote, smaller, derive, lemma, property, lower, bounded, assume, prove, element, order, introduce] [average, function, based, cone, standard] [number] [video, frame, figure, ieee, outperforms, applied, compare, larger, snr, restricted] [recovery, conventional, sparse, program, convex, problem, presence, measurement, rip, mse, theoretical, recover, exact, reconstruction, reconstructed, csk, recovering, min, rnk, application, stable, random, solving, matrix, generalized, suppose, sensing, support, compressive, noiseless, satisfy, compressed, propose] [dynamic, absence, signal] [linear, error, real, vector, table, data, select, generalize, row]
Efficient Reinforcement Learning for High Dimensional Linear Quadratic Systems
Morteza Ibrahimi, Adel Javanmard, Benjamin Van Roy


[algorithm, regret, set, theorem, bound, probability, lemma, proof, result, assume, order, denote, consider, prove, hold, corollary, asymptotic, extended, theory, exists, case, marketing, bounded, suitable, introducing] [control, cost, optimal, cmin, state, episode, controller, cumulative, advertising, average, hjs, unknown, regulator, initial, expected, max] [parameter, number, model, event, log] [work, ieee, spatial, second] [high, dimensional, estimated, matrix, sparse, problem, stochastic, solution, quadratic, optimization, proposition, condition, recovery, estimate, decomposition, stanford, corresponding] [system, time, adaptive, controlled, dynamic, principle, equation] [linear, learning, row, logarithmic, constraint]
Ancestor Sampling for Particle Gibbs
Fredrik Lindsten, Michael Jordan, Thomas Schön


[set, probability, path, consider, mixing, assume, statistical, note, sequence, explicit, complexity, order, series] [simulation, state, sequential, idea, trajectory, based, approach] [backward, sampling, particle, truncation, ancestor, smc, sampler, gibbs, inference, gaussian, carlo, monte, model, draw, process, sample, will, mcmc, pmcmc, conditional, computational, forward, ssms, bayesian, compute, markov, degenerate, university, number, ancestral, proposal, csmc, chain, adaption, tree] [level, figure, apply, reason, target, accuracy] [density, method, step, problem, journal, procedure, running, application, random, strong] [smoothing, time, increase, system, noise, nonlinear] [approximation, linear, proposed, address, error, construct, kernel]
Modelling Reciprocating Relationships with Hawkes Processes
Charles Blundell, Katherine Heller, Jeff Beck


[probability, measure, set, consider, interesting, statistical] [people, based, total, intelligence, behaviour, future, approach, uncertainty] [model, hawkes, process, poisson, social, cluster, gpq, npq, irm, number, enron, bob, nuv, structure, conv, event, interaction, nqp, mallory, will, inference, relational, email, log, university, pair, david, markov, military, reciprocity, whilst, reciprocating, infer, conversation, predictive, bayesian, thinning, alice, full, latent, parameter, inferred, modelled, involving, partition, declared, modelling] [figure, work, learn, structured] [rate, mid, edge, graph, random, journal, denotes, stochastic, relationship] [time, group] [data, point, table, learning, counting, synthetic]
Expectation Propagation in Gaussian Process Dynamical Systems
Marc Deisenroth, Shakir Mohamed


[true, algorithm, set] [state, approximate, function, expectation, based, standard, transition, average, uncertainty, international, conference, required, updated] [gaussian, inference, posterior, distribution, latent, message, process, log, bayesian, gpdss, gpads, propagation, factor, model, update, linearization, marginal, moment, cavity, predictive, computing, implicit, gpds, partition, inferred, compute, kalman, nllx, backward, forward, form] [capture, matching, directly, compared] [measurement, covariance, general, passing, corresponding, iterative] [dynamical, smoothing, time, nonlinear, motion, account, system, computed, prediction, noise] [data, approximation, test, learning, existing, improved, performance, projection, machine, accurate, synthetic]
A quasi-Newton proximal splitting method
Stephen Becker, Jalal Fadili


[algorithm, active, set, case, theorem, example, class, corollary, denote, assume, consider, version, term, result, piecewise, remark, extended, loss, strictly] [function, standard, approach, cost, continuous] [log, compute, update] [simple, root, second, previous] [method, convex, gradient, diagonal, proximity, optimization, newton, proxv, solve, proximal, inverse, constrained, problem, splitting, siam, general, unique, exact, positive, hessian, min, generalized, unconstrained, step, quadratic, derivative, objective, lasso, descent, numerical, analysis, convergence, smooth, proxhi, scalar, spectral, asa, scaled, lipschitz, solving, semismooth, symmetric, sparse, constant] [] [operator, projected, approximation, machine, separable, well, proposed, sorting]
Minimizing Uncertainty in Pipelines
Nilesh Dalvi, Aditya Parameswaran, Vibhor Rastogi


[set, probability, consider, query, hard, best, theorem, denote, minimizes, dag, complexity, active, pick, picking, report, algorithm, example, term, result, extended, note, closest, pipeline, ree, correctness, minimize, version, extractor, john, polynomial, proof, class, alina, label] [uncertainty, expected, approximate, correct, total, goal, intractable, function, optimal, based] [node, compute, tree, probabilistic, conditioned, will, model, ancestor, number, directed, inference, event, expression] [leaf, work, single, second, root, simple, computer] [problem, graph, technical, incremental, general, objective, denotes, size, weighted, graphical] [output, time] [data, reduction, approximation, linear, learning, operator, item]
Mandatory Leaf Node Prediction in Hierarchical Multilabel Classification
Wei Bi, James Kwok


[multilabel, algorithm, label, mlnp, set, probability, supernode, loss, unassigned, nap, masr, snv, satisfying, hbr, dag, best, extended, called, multiclass, query, path, example, minimizes, mandatory, partial, smaller, pheno] [hierarchy, conference, international, optimal, function, search, approach, water] [hierarchical, node, number, tree, posterior, will, hmc, likelihood, updating, log, parent, bayesian, enron] [leaf, root, figure, performed, pattern, training, internal] [proposition, problem, composite, assumption, symmetric, pij, method, gene, positive, greedy, solution] [prediction, time, threshold, processing] [data, learning, proposed, labeled, large, machine, table, largest, test, existing]
Efficient Monte Carlo Counterfactual Regret Minimization in Games with Many Player Actions
Richard Gibson, Neil Burch, Marc Lanctot, Duane Szafron


[regret, probability, set, theorem, bound, algorithm, best, bounded, consider, prove, provide, note] [player, average, game, action, strategy, counterfactual, exploitability, mccfr, sampled, cfr, cumulative, visited, abstract, chance, faster, poker, current, michael, equilibrium, betting, opponent, nash, return, dice, vanilla, tighter, converges, extensive, history, required, starting, round, max, martin, indicate, play, terminal] [sampling, number, sample, computing, tree, monte, carlo, parameter] [previous, figure, single, sum, recall, work] [minimization, random, iteration, solution, convergence, size] [response, time] [averaged, data, well]
Multilabel Classification using Bayesian Compressed Sensing
Ashish Kapoor, Raajay Viswanathan, Prateek Jain


[active, label, multilabel, set, note, case, consider, fraction] [uncertainty, standard, based, key, total, function] [model, inference, number, bayesian, posterior, latent, gaussian, update, distribution, probabilistic, observed, prior, variational, joint, gamma, fully, process, turn] [precision, training, figure, accuracy, unlabeled, baseline, handle, considering, obtains] [method, compressed, sensing, random, dimensional, procedure, compressive, jointly, positive, sparse] [output, framework, traditional, prediction, weight, input] [learning, data, svm, missing, proposed, space, point, regression, vector, test, unobserved, machine, labeled, naturally, dataset, large, selection, select, informative, datasets, selected, criterion, better, linear]
Scaling MPE Inference for Constrained Continuous Markov Random Fields with Consensus Optimization
Stephen Bach, Matthias Broecheler, Lise Getoor, Dianne O'Leary


[algorithm, derive, set, example, case, feasible, statistical, consider, introduce, class, refer] [continuous, function, approach, average, based, evaluated, conference] [social, number, update, inference, probabilistic, markov, modeling, model, discrete, component, will] [figure, work, single] [problem, mpe, optimization, consensus, method, random, arg, ccmrfs, running, objective, solving, constrained, subproblems, solve, ccmrf, graphical, broecheler, iterative, college, dual, solved, min, scalability, xxt, naive, copy, procedure, edge, clipping, convex] [potential, time, linearly, framework, intrinsic, represent, speed, direction] [constraint, linear, learning, point, vector, machine, superior, relative]
Query Complexity of Derivative-Free Optimization
Kevin Jamieson, Rob Nowak, Ben Recht


[oracle, comparison, algorithm, bound, lower, inf, pairwise, theorem, dfo, denote, case, probability, class, set, upper, lemma, respect, sequence, setting, assume, cfi, active, fact, bandit, query, proof, minimum, complexity, bounded, implies, depend, prove, uniformly, divergence, note, considered] [function, search, evaluation, optimal, max, based, correct, expectation] [log, distribution, will, model, parameter, gaussian, number, exp] [simple, noisy, apply, work, human] [optimization, convex, stochastic, random, lipschitz, convergence, rate, gradient, problem, procedure, journal, constant, objective, arg, independent, robust, analysis] [noise, direction] [learning, well, small, dimension]
A Stochastic Gradient Method with an Exponential Convergence _Rate for Finite Training Sets
Nicolas Le Roux, Mark Schmidt, Francis Bach


[testing, set, loss, algorithm, logistic, case, note, respect, paper, sequence, assume] [cost, optimal, function, faster, average, standard, based, converges, additional, deterministic] [form, number, full, supplementary, exponential] [training, previous, work, larger, despite, improve] [convergence, method, rate, step, sag, gradient, size, stochastic, constant, iteration, proposition, achieve, iag, analysis, afg, averaging, convex, problem, independent, journal, incremental, accelerated, fik, siam, pass, pegasos, rda, optimization, momentum, nosg, esg, practical, steepest, assumption, objective, smooth, storage, regularized, iterates, afford] [effective, basic] [data, linear, learning, large, machine, better, approximation, test, selection, regularization]
Emergence of Object-Selective Features in Unsupervised Feature Learning
Adam Coates, Andrej Karpathy, Andrew Ng


[algorithm, best, class, set] [conference, international, key] [will, scalable, distribution, model, form, full, highly, structure] [simple, complex, figure, invariant, object, feature, learned, layer, train, unlabeled, pooling, unsupervised, learn, invariance, scale, image, training, trained, work, human, discover, second, pixel, visualize, face, combine, youtube, activate, local, directly, randomly, agglomerative, clear, applying, computer, identify, single, commonly, supervision] [procedure, random, achieve, sparse, density, method, type] [cell, selective, group, system, neural, coding] [learning, linear, labeled, clustering, large, data, dataset, construct, extremely, existing, small, machine, better]
Learning as MAP Inference in Discrete Graphical Models
Xianghang Liu, James Petterson, Tiberio Caetano


[loss, label, set, algorithm, runtime, note, paper, example, complexity, statistical, main, probability, fact, logistic, argument, margin] [continuous, good, approach, open, option, key, observation, function, approximate] [discrete, inference, model, will, number, parameter, ability, discretization, fully, chain] [map, work, training, accuracy, structured, figure, directly, complex, svms] [random, problem, convex, graph, optimization, formulation, high, graphical, assumption, method, clique, exact, robustness, theoretical, general, noiseless, analysis] [framework, connected, noise, time, direction] [learning, linear, machine, binary, svm, small, data, well, vector, datasets, regularization, space, regularizer]
Learning curves for multi-task Gaussian process regression
Peter Sollich, Simon Ashton


[bayes, case, asymptotic, consider, lower, example, order, square, set, asymptotically, setting, assume, discussion, term] [average, function, approach, initial, good] [number, gaussian, will, form, posterior, process, pure, fully, prior, latent, explicitly, expression] [training, second, figure, feature, single, level, sum, learned, product] [matrix, covariance, smooth, block, diagonal, corresponding, general, solving] [curve, noise, prediction, input, decay, transfer, gain, limit, study, effective, solid, surface] [learning, task, error, regression, approximation, mit, large, kernel, variance, machine, linear, averaged, small, london, space, rough, individual]
A mechanistic model of early sensory processing based on subtracting sparse representations
Shaul Druckmann, Tao Hu, Dmitri Chklovskii


[bregman, algorithm, series, result, consistent, note, assume, consider] [function] [predictive, transmission, will, number, model] [feature, representation, view, higher, internal, considering, cross, single] [sparse, step, iteration, bandwidth, journal, solution, analysis, convex] [stimulus, principal, feedback, time, circuit, activity, coding, interneurons, sensory, negative, interneuron, response, early, neuron, neural, threshold, linearized, equation, prediction, brain, processing, transient, transmitted, olfactory, transmit, ganglion, input, natural, direct, channel, inhibitory, membrane, nature, longer, framework, signal, output, receptive, nonlinear, residual, implement, delay, synaptic, efficient, retinal, lateral, inhibition, noise, defined, network, correlated, limited, integration] [linear, projection, accurate, common, point, error, well]
Multi-Stage Multi-Task Feature Learning
Pinghua Gong, Jieping Ye, Changshui Zhang


[algorithm, bound, lemma, set, theorem, remark, proof, inequality, empirical, comparison, result, case, loss, denote, upper, annals, order, term] [max, based, state] [parameter, gaussian, number, exp] [feature, training, mri, side, multiple, figure] [estimation, sparse, analysis, condition, min, formulation, matrix, assumption, solution, theoretical, propose, random, eigenvalue, lasso, convex, nonzero, zhang, regularized, incoherence, denoted, achieves] [stage, weight, effectiveness, nmse, noise] [learning, error, msmtfl, data, proposed, averaged, common, task, shared, row, performance, regularization, synthetic, ratio, jalali, well, dirty, large, paramter, isolet, smallest, demonstrate, restrictive, dirtymtl, weaker, amse]
An Integer Optimization Approach to Associative Classification
Allison Chang, Dimitris Bertsimas, Cynthia Rudin


[transaction, class, iter, mio, set, list, orc, highest, algorithm, win, mining, ranked, frontier, note, rgen, bound, board, interpretability, upper, term, lim, ranking, paper, integer, interestingness, tij, applicable, chose, report, called] [decision, optimal, average, cart, international, null, based, conference, add, approach] [number, model, generate, parameter] [accuracy, train, figure, work, training, include, second, designed, multiple, interpretable, top, build, corresponds, higher, body, help] [rank, formulation, objective, optimization, solve, support, method, achieve, scaled, correctly] [rule, association, time, predicts] [test, data, constraint, table, learning, machine, regularization, vector, point, dataset, small, binary]
Distributed Non-Stochastic Experts
Varun Kanade, Zhenming Liu, Bozidar Radunovic


[payoff, regret, communication, algorithm, coordinator, consider, case, lower, bound, fri, fpl, oblivious, adversary, query, counter, online, dfpl, received, follow, note, theorem, leader, setting, assume, simultaneously, proof, lemma, asymptotically, observe, decide, guarantee, maintain, probability, best, forecaster, perturbed, applicable, upper, version, lef, choose, term] [expert, cumulative, expected, receives, total, cost, optimal, approximate, standard, required, playing] [site, model, number, will, parameter, described, variable, full] [work, simple, figure] [distributed, block, random, problem, achieves, step, regularized, optimization, stochastic] [time, prediction, system, phase, noise, natural] [vector, better, learning, large, machine]
Latent Graphical Model Selection: Efficient Methods for Locally Tree-like Graphs
Anima Anandkumar, Ragupathyraj Valluvan


[complexity, set, consider, minimum, provide, class, denote, algorithm, lower, probability, bound, fraction, annals, chosen, note, depend] [based, tradeoff] [latent, model, tree, observed, number, hidden, node, ising, sample, loopy, lda, topic, girth, log, structure, maximum, parameter, markov, perplexity, degree, dmax, tractable, variable, bic, distribution, discrete, process, gaussian, discovering, pmi] [local, depth, work, incorporate, unlabeled, evaluate] [method, graph, graphical, edge, correlation, estimation, random, dmin, establish, presence, require, reconstruction] [decay, potential, regime, natural, long, output] [learning, proposed, distance, locally, performance, selection, large, machine, data, binary]
Near-optimal Differentially Private Principal Components
Kamalika Chaudhuri, Anand Sarwate, Kaushik Sinha


[algorithm, complexity, set, lower, theorem, case, theory, gap, guarantee, interesting, bound, empirical, chose, change, induced] [utility, approximate, required, sensitive, standard, future] [sample, distribution, parameter, exponential, number, sampling, log, gaussian] [ppca, work, top, second, single, accuracy, level] [matrix, pca, random, method, subspace, general, theoretical, perturbation, low, procedure, eigenvector, scaling, rank, singular, eigenvalue] [output, noise, mechanism, computation] [data, privacy, private, differentially, sulq, approximation, differential, dimension, performance, sensitivity, well, talwar, reduction, nonprivate, dimensionality, orthonormal, special, dwork, variance, point, real, large, vector, machine, bingham]
Random function priors for exchangeable arrays with applications to graphs and relational data
James Lloyd, Peter Orbanz, Zoubin Ghahramani, Daniel Roy


[theorem, set, case, probability, measurable, statistical, consider, supported, uniform, suitable, discussion] [function, school, sampled, continuous, international, conference] [model, latent, gaussian, array, exchangeable, process, bayesian, relational, prior, xij, nonparametric, sampling, distribution, number, observed, uij, form, variable, inference, parameter, conditional, slice, gplvm, structure, eigenmodel, xji, modeling, posterior, modelling, sample, university, probabilistic, exchangeability] [representation, protein, scale, work] [random, graph, matrix, journal, sparse, general, assumption, stochastic, high, symmetric, adjacency] [wij, network, processing, neural, natural, noise, interpreted] [data, space, kernel, machine, learning, pmf, approximation, common]
Mirror Descent Meets Fixed Share (and feels no regret)
Nicolò Cesa-Bianchi, Pierre Gaillard, Gabor Lugosi, Gilles Stoltz


[regret, loss, online, forecaster, sequence, algorithm, theorem, shifting, bound, share, dtv, best, corollary, notion, lemma, case, setting, proof, comparison, mirror, chosen, measure, note, simplex, consider, order, quantity, minimize, introduced, compete, paris, mixing, element, introduce] [discounted, expert, focus, cumulative, horizon, shift, standard, conference, function, round, current, total, chooses, goal, max] [update, number, supplementary, form] [tracking, corresponds, sum, work] [generalized, convex, analysis, general, optimization, descent, suppose, journal, satisfy] [adaptive, tuning, prediction, regularity, time, difference, expressed, ecole] [learning, small, machine, special, shared, linear, projection, projected, generalizes, logarithmic]
Perfect Dimensionality Recovery by Variational Bayesian PCA
Shinichi Nakajima, Ryota Tomioka, Masashi Sugiyama, S. Derin Babacan


[theorem, bayes, true, result, energy, assume, derived, derive, note, property, probability, set, empirical, situation, implies, algorithm] [function, observation, based, search] [bayesian, posterior, distribution, variational, observed, model, probabilistic, log, sample, parameter, analytic, gaussian] [global, simple, depicted, local, figure, automatic, second, perform] [evb, matrix, solution, estimator, sparsity, condition, recovery, singular, pca, estimation, free, random, rigorous, theoretical, rank, practical, mpul, analysis, shrinkage, independence, rate, covariance, journal, proposition, estimated, tokyo, cah, minimizer, quartic, denotes, numerical, cbh, exact, vbmf] [noise, threshold, perfect, implementation, behavior, limit, principal, theoretically] [variance, dimensionality, selection, factorization, large, well, approximation, advantage]
Spectral learning of linear dynamics from generalised-linear observations with application to neural population data
Lars Buesing, Jakob Macke, Maneesh Sahani


[set, algorithm, consistent, true, case] [observation, function, state, approach, initial, approximate, based] [model, observed, parameter, gaussian, latent, process, moment, distribution, computational, likelihood, factor, poor, generated, inference, poisson, number, described, loading] [training, truth, multiple, ground, local, compared] [matrix, hankel, size, covariance, method, estimation, estimate, random, subspace, underlying, spectral, singular, yield] [pldsid, neural, ssid, time, initialisation, dynamical, plds, spectrum, system, external, initialised, instantaneous, difference, conversion, input, activity, computed] [data, performance, linear, well, better, large, approximation, learning]
Convex Multi-view Subspace Learning
Martha White, Yaoliang Yu, Xinhua Zhang, Dale Schuurmans


[set, loss, algorithm, lemma, runtime, note, case, theorem, increasing, appendix] [optimal, max, approach, key, goal] [latent, number, conditional, variable, allows, conditionally] [training, representation, figure, denoising, local, image, global, noisy, higher, face, achieved, single, canonical, compare] [msl, subspace, convex, ssl, min, lsl, matrix, norm, yte, formulation, reconstruction, problem, xte, recovery, independence, optimization, sparse, recovering, rank, objective, independent, analysis, solve, proposition, lighting, concave, runtimes, constrained, dual, recover, alternating, hte, correlation, zktr, dimensional, procedure, minimization, solution, high] [noise, expressed, developed] [learning, data, error, synthetic, quality, dimensionality, shared, boosting, ratio, cca, dataset, inner, worse, linear]
Regularized Off-Policy TD-Learning
Bo Liu, Sridhar Mahadevan, Ji Liu


[algorithm, set, mirror, term, loss, comparison, design, result, note, online] [function, tdc, based, conference, mspbe, approach, reinforcement, state, international, bellman, policy, eligibility, action, formulated, cost, max, inverted, transition, concern, pendulum, reward, approximate, star] [number, compute, markov, sample, sampling, computing] [feature, figure, temporal, experiment, learn, representation] [convex, gradient, regularized, objective, problem, formulation, optimization, convergent, descent, analysis, proximal, convergence, stochastic, method, sparse, subgradient, matrix, novel, condition, minimization, min, underlying] [equation, computation, basis, difference, presented, material, experimental] [linear, learning, selection, regularization, machine, proposed, error, table, space, performance, well, column]
Multi-criteria Anomaly Detection using Pareto Depth Analysis
Ko-Jen Hsiao, Kevin Xu, Jeff Calder, Alfred Hero


[set, best, theorem, denote, algorithm, strictly, consider, measure] [pareto, anomaly, pda, front, dyad, anomalous, nominal, approach, grid, hero, trajectory, optimality, create, located, based, conference, scalarization, search, iii] [sample, will, number, distribution, associated, supplementary, computing] [detection, multiple, training, score, depth, figure, single, detect, experiment, view, corresponds, compared, second, sum, work] [corresponding, method, problem, analysis, density, positive] [time, weight, processing, study] [linear, test, data, dissimilarity, learning, nearest, proposed, criterion, combination, item, auc, machine, performance, dij, large, better, geometry]
Synchronization can Control Regularization in Neural Systems via Correlated Noise Processes
Jake Bouvrie, Jean-Jacques Slotine


[consider, term, theorem, respect, assume, redundancy, set, case] [observation, equilibrium, trajectory, decision, control, cost] [will, process, computational, form, model, gaussian, distribution, collection, degree, prior, parameter] [noisy, impact] [gradient, convergence, solution, stochastic, problem, independent, descent, rate, general, averaging, plot, quadratic, covariance, random, distributed] [noise, synchronization, coupling, network, neural, system, time, strength, coupled, correlated, mechanism, visual, equation, corrupted, observational, sensory, nervous, behavior, ambient, role, brownian, adjust, dwt, experimental, dynamic, difference, brain, increase, adaptive, plausible, implementing] [regularization, learning, error, parallel, variance, regression, approximation, data, common, functional, small]
Calibrated Elastic Regularization in Matrix Completion
Tingni Sun, Cun-Hui Zhang


[theorem, bound, calibrated, algorithm, proof, inequality, lemma, smaller, order, proper, iid, main, uniformly, term] [optimal, simulation, unknown] [log, sample, factor, hpt, observed, rest, allows] [level, training, second, noisy, target, svd, sum, figure] [matrix, size, rank, lasso, coherence, norm, penalty, frobenius, singular, completion, estimator, exact, nuclear, requirement, problem, recovery, iterative, arg, min, random, analysis, elastic, suppose, estimation, convex, noiseless, minimization, quadratic, proposition, solution, estimated, method, technical, distributed, scaled, simpler] [spectrum, noise, proportion, complete, requires] [error, linear, proposed, data, large, space, ratio, regularization, parallel, learning, test, constraint, advantage, machine]
Stochastic Gradient Descent with Only One Projection
Mehrdad Mahdavi, Tianbao Yang, Rong Jin, Shenghuo Zhu, Jinfeng Yi


[algorithm, lemma, theorem, bounded, online, bound, inequality, assume, regret, proof, probability, main, note, set, sequence, convexity, order, term, feasible, result, exists, depend, remark, complexity] [function, optimal, domain, standard, max, key, idea, cost] [variable, exp, parameter, updating, will, develop, computational, update] [sgd, complex, work] [convex, optimization, stochastic, gradient, rate, convergence, solution, general, problem, descent, min, step, assumption, iteration, objective, analysis, norm, achieve, dual, primal, method, high, eigenvalue, lagrangian, intermediate, achieves, greedy, strong, sparse, mild, usa, positive, theoretical] [making, requires, computation] [projection, proposed, learning, constraint, machine, approximation, projecting, unbiased, psd]
Multiclass Learning Approaches: A Theoretical Comparison with Implications
Amit Daniely, Sivan Sabato, Shai Shalev Shwartz


[hypothesis, class, multiclass, msvm, complexity, ecoc, theorem, halfspaces, label, analyze, erm, set, probability, note, lemma, respect, case, lower, bound, denote, hova, algorithm, corollary, upper, regret, allwein, crammer, chosen, main, daniely, consider, proof, stated, length, singer, exists, result, best, realizable, roughly] [choice, based, function, approach] [sample, tree, will, distribution, number, associated, rest] [training, randomly, larger, perform] [estimation, random, analysis, underlying, method, problem, essentially, journal, graph, high, matrix, constant, close] [prediction, output, left, difference] [error, approximation, binary, code, learning, dimension, large, reduction, vector, small, better, linear, distance, performance]
Fusion with Diffusion for Robust Visual Tracking
Yu Zhou, Xiang Bai, Wenyu Liu, Longin Jan Latecki


[set, measure, algorithm, lim, lemma, best, complexity, consider, probability] [based, conference, approach] [diffusion, process, compute, indicates, number, pair, will, represented] [tracking, frame, computer, ivt, ieee, single, pattern, precision, target, location, tpg, center, fusion, product, multiple, image, vision, vec, combine, simple, three, patch, improve, appearance, object, matching, society, background, color, tracker, coke, cliff, surfer] [graph, robust, matrix, method, weighted, iterative, analysis] [visual, threshold, time, experimental, weight] [similarity, linear, learning, error, machine, data, proposed, table, tensor, sylvester, task, candidate, column, original, vector, challenging]
Learning Probability Measures with respect to Optimal Transport Metrics
Guillermo Canas, Lorenzo Rosasco


[measure, probability, empirical, bound, order, lemma, lower, upper, set, theorem, voronoi, inequality, note, supported, algorithm, consider, ewp, closest, absolutely, quantizer, derived, series, wasserstein, transport, borel, induced, instance, best, prove, proof, sense, term, theory, drawn, respect, bounded, minimizing, measurable, springer] [optimal, cost, continuous, expected] [probabilistic, sample, distribution, described, concentration, discrete] [unsupervised, matching, weak, applied] [convergence, problem, rate, decomposition, size, close, density, support, random, journal, der] [population, connection, transportation, equation, behavior, output, arrow, study, unit] [learning, distance, quantization, performance, existing, space, metric, manifold, large, approximation, riemannian, point, well, projection]
Value Pursuit Iteration
Amir massoud Farahmand, Doina Precup


[algorithm, set, probability, sequence, empirical, bound, upper, measurable, provide, main, theorem, result, denote, inf, absolutely, consider, follow, case, theory, proof, paper] [function, vpi, optimal, policy, qmax, omp, bellman, future, reinforcement, conference, behaviour, transition, optimality, approximate, international, ronald, goal, based, csaba, deterministic, earlier, barron, farahmand, initial] [number, distribution, model, link, will, supplementary, atom, markov, allows, nonparametric] [representation, learned, feature, second, matching, product] [dictionary, iteration, sparse, norm, pursuit, greedy, close, theoretical, problem, procedure, analysis, method, estimate, size] [represent, study, basis] [approximation, error, learning, space, selection, operator, machine, large, orthogonal, linear]
"Compressive neural representation of sparse, high-dimensional probabilities"
Xaq Pitkow


[probability, perceptron, set, theory, bounded] [function, state] [distribution, number, posterior, process, hidden, will, probabilistic, sampling, inference, generated, structure, represented] [figure, boltzmann, pattern, noisy, reconstruct, representation, ieee, visible, randomly, simple, product] [compressive, random, sparse, compressed, sensing, measurement, matrix, reconstruction, cxx, arxiv, independent, nonzero, preserved, reconstructed] [signal, sgn, perceptrons, represent, nonlinear, neural, brain, activity, evidence, input, population, threshold, feedforward, ensure, wij, natural, sensory, encoding, accurately] [binary, large, inner, vector, space, variance, error, linear, dimensionality, small, embedding]
Learning Manifolds with K-Means and K-Flats
Guillermo Canas, Tomaso Poggio, Lorenzo Rosasco


[set, case, voronoi, empirical, order, setting, measure, note, bound, algorithm, consider, example, approximated, probability, quantity, respect, piecewise, theorem, risk, provide, fundamental, statistical, instance, class, considered, result, mathematical, symposium, depend, supported, studied, minimum, sense, best] [expected, optimal, international] [distribution, sample, discrete, form, typically] [global, embedded, ieee, region, higher, curvature] [reconstruction, analysis, problem, solution, assumption, high, constant, dictionary, geometric, size, dimensional, general, technical, random] [behavior, study, framework, unit, equation] [approximation, manifold, space, learning, error, data, hilbert, quantization, well, dimensionality, tangent, performance, dimension, large, clustering, orthogonal, better, naturally]
A Polynomial-time Form of Robust Regression
Yaoliang Yu, Ozlem Aslan, Dale Schuurmans


[loss, bounded, set, empirical, consistency, appendix, consider, note, risk, denote, polynomial, consistent, guarantee, probability, theorem, bound, assume, case] [standard, function, based, unknown, strategy, evaluation, approximate, generally, deterministic, remains, allow, domain] [parameter, form, computational, will, number, base, variational, sample, allows] [single, training, local, weak, compared, experiment, leverage, randomly] [convex, robust, min, estimator, robustness, general, method, minimization, breakdown, arg, outlier, solution, rounded, tikhonov, random, regularized, relaxation, lts, problem, unbounded, relaxed, alternating, tractability, estimate, objective, achieve, establish, trimmed, conducted] [response, expressed, experimental, adaptive] [regression, data, regularization, vector, approximation, test, kernel, learning, machine, point, proposed, table]
Iterative ranking from pair-wise comparisons
Sahand Negahban, Sewoong Oh, Devavrat Shah


[algorithm, walk, probability, btl, lemma, ranking, theorem, main, pairwise, result, bound, assume, induced, consider, proof, comparison, version, provide, aji, denote, setting, chosen, term, outcome] [stationary, transition, based, aggregation, choice, state, player] [distribution, model, will, markov, chain, number, pair, log, ordering, sampling, reversible, sample, dmax, analytic, count, likelihood] [score, work, object, recall] [random, matrix, rank, graph, spectral, aij, underlying, pij, analysis, positive, centrality, iterative, problem, estimate, theoretical, constant, corresponding, high, rate, estimated, establish, general] [presented, ideal, preferred, desired, weight, long, natural, wide] [error, item, well, vector, data, performance, operator, ratio]
Identifiability and Unmixing of Latent Parse Trees
Daniel Hsu, Sham Kakade, Percy Liang


[def, algorithm, consider, sentence, mixing, probability, note, set, exists, theorem, varies, iff, length, consists, measure, class, integer] [parsing, observation, function, parse, based, state, goal, standard, additional] [model, tree, dependency, observed, constituency, latent, topology, parameter, distribution, compound, pcfg, will, heck, markov, node, hidden, generate, number, unmixing, compute, structure, word, generative, probabilistic, mixture, collection, turn, joint, dentifiability] [figure, jacobian, corresponds, sum, unsupervised, applied, work, root, product, restricted, local] [matrix, recover, spectral, estimate, diagonal, rank, numerical, problem, estimation, general] [left, system] [learning, space, binary, row, technique]
Persistent Homology for Learning Densities with Bounded Support
Florian Pokorny, Carl Henrik Ek, Hedvig Kjellström, Danica Kragic


[probability, consider, bounded, interval, set, example, order, note, class, subset] [approach, based, max, choice, admissible, standard] [sample, gaussian, model, mixture, compute, number, sampling, probabilistic, topology, prior, will, structure, parameter, inferred] [figure, region, top, complex, display, work, track] [density, topological, bandwidth, homology, support, corresponding, amise, betti, estimator, racetrack, barcode, estimation, supp, mid, min, topologically, method, size, barcodes, norm] [persistent, spherical, displayed, position, connected, velocity, study, indicated, computed, ith, motion] [data, kernel, learning, space, dimension, large, vector, error, synthetic]
3D Social Saliency from Head-mounted Cameras
Hyun Soo Park, eakta Jain, Yaser Sheikh


[algorithm, respect, calibration] [gaze, ray, primary, head, people, eye, regard, apex, concurrence, oblique, person, understanding, plane, based, current, standard, cone, musical, move, third] [social, distribution, model, number, form, will, gaussian, maximum, structure] [camera, figure, view, center, saliency, scene, multiple, capture, pose, top, reconstruct, ground, ieee, truth, face, side, representation, directly, computer, human, location, image, variation, half] [estimate, density, method, estimation, estimated, reconstructed, gradient, coordinate, estimating] [orientation, motion, direction, group, position, system, visual, equation, unit] [point, distance, vector, geometry, projection, real, kernel]
Mixing Properties of Conditional Markov Chains with Unbounded Feature Functions
Mathieu Sinn, Bei Chen


[theorem, lim, mixing, sequence, ergodicity, denote, note, strictly, asymptotic, proof, result, order, paper, introduce, theory, ergodic, consider, write, bound, set, exists, consistency, extended, assume, hold, measurable, version, establishes, probability, cmcs, fundamental, satisfying, introduced, annals, corollary, statistical] [function, key, transition, conference, international, behaviour, context, tail, max] [conditional, markov, distribution, likelihood, model, hidden, will, maximum, concentration, component, marginal] [feature, aligned, second] [random, suppose, unbounded, condition, assumption, establish, analysis, step, proposition, hessian, constant, established, positive] [limit, time, potential, central, difference, study] [space, vector, learning, linear, machine]
Multiple Choice Learning: Learning to Produce Multiple Structured Outputs
Abner Guzmán-Rivera, Dhruv Batra, Pushmeet Kohli


[set, algorithm, loss, best, energy, notice, observe, minimizes, pairwise, label] [function, goal, standard] [model, number, typically, parameter, latent] [multiple, single, structured, segmentation, produce, training, map, learn, score, image, protein, work, pixel, train, figure, baseline, outperforms, object] [min, objective, problem, descent, formulation, coordinate, optimization, graph, positive, iteration, step] [prediction, output, input, natural, plausible, producing] [mcl, error, learning, ssvm, diverse, predictor, machine, interactive, task, space, quality, predicting, test, ensemble, performance, proposed, accurate, clustering, mapping, dataset, user]
Forward-Backward Activation Algorithm for Hierarchical Hidden Markov Models
Kei Wakabayashi, Takao Miura


[probability, algorithm, path, sequence, set, complexity, notation] [state, transition, execution, null, based, function, calculate, observation, terminates] [markov, hhmm, hidden, hierarchical, parameter, inference, sampling, hhmms, chain, gibbs, forward, backward, minsr, fba, ftd, latent, adij, tree, structure, hmms, log, distribution, likelihood, gbiv, ffb, enables, gadij, termination, fbas, flattening, initialization, model, sampler, identical, biv, conditional] [level, single, figure, unsupervised, dbn, corresponds, depth, apply, training] [method, estimation, convergence, iteration] [activation, time, absolute, direct, dynamic] [learning, proposed, space, table, existing, conduct, data, constraint]
Communication-Efficient Algorithms for Statistical Optimization
Yuchen Zhang, John Duchi, Martin Wainwright


[algorithm, empirical, theorem, statistical, risk, provide, set, denote, order, paper, loss, bound, corollary, main, smoothness, term, probability, best, advertisement] [average, standard, based, optimal, approach, function, search] [number, parameter, mixture, distribution, sample, form, log] [figure, single, local, second, training] [bavgm, avgm, distributed, stochastic, method, estimate, gradient, convex, convergence, assumption, problem, rate, averaging, descent, size, solving, optimization, versus, plot, estimator, matrix, require, hessian, constant, achieves, procedure, analysis, journal, minimization] [bias, population, processing, negative, prediction] [data, dataset, bootstrap, vector, error, regression, linear, performance, machine, large, approximation, better, user]
Approximate Message Passing with Consistent Parameter Estimation and Applications to Sparse Learning
Ulugbek Kamilov, Sundeep Rangan, Alyson Fletcher, MIchael Unser


[algorithm, asymptotic, consider, true, lim, result, class, set, consistency, oracle, theorem, paper, asymptotically, sequence, consistent, suitable] [approximate, max, function, correct, unknown, expectation, state, choice, belief, deterministic] [parameter, gaussian, message, will, distribution, prior, joint, described, bayesian, model, propagation, conditional, variable] [ieee, figure, simple] [gamp, estimation, random, method, scalar, sparse, componentwise, compressed, measurement, mse, reconstruction, passing, amp, estimate, converge, evolution, lasso, general, analysis, matrix, assumption, sensing, problem, family] [adaptive, adaptation, signal, output, input, noise, channel, behavior, considerable, nonlinear] [large, linear, vector, performance, learning, variance, ratio]
Feature Clustering for Accelerating Parallel Coordinate Descent
Chad Scherrer, Ambuj Tewari, Mahantesh Halappanavar, David Haglin


[algorithm, loss, upper, note, order, consider, case, design, smaller, choose, set, radius, bound, provide] [function, heuristic, expected, based, updated, explore] [number, will, parameter] [feature, figure, single, work, product, second, simple, three, ten] [coordinate, descent, convergence, analysis, block, clustered, greedy, regularized, spectral, convex, corresponding, objective, step, rel, stochastic, norm, family, general, nonzero, minimization, processor, smooth, proposition, solution, rate] [time, absolute, weight, memory] [clustering, parallel, randomized, data, nonzeros, inner, nnz, large, largest, table, scherrer, regularization, machine, special, shotgun, select, euters, thread, separable, kdda, national, datasets, gathered, accept]
Multiple Operator-valued Kernel Learning
Hachem Kadri, Alain Rakotomamonjy, Francis Bach, Philippe Preux


[algorithm, consider, case, set, respect, setting, order] [function, based, continuous] [number, integral, associated] [multiple, sum, ecog, work] [problem, optimization, positive, block, method, matrix, solving, min, solve, dual, descent, solution, analysis, norm, convergence, propose, estimating, lagrangian, coordinate] [movement, output, response, system, prediction, framework, amplitude, bci, time, nonlinear, residual, moving, interface, input, signal] [kernel, functional, learning, operator, regression, movkl, linear, hilbert, combination, data, space, reproducing, ridge, vector, real, krr, lagrange, dataset, rkhs, multiplication, proposed, mkl, rsse, error]
Finding Exemplars from Pairwise Dissimilarities via Simultaneous Sparse Recovery
Ehsan Elhamifar, Guillermo Sapiro, Rene Vidal


[algorithm, consider, pairwise, subset, probability, sentence, note, set, smaller, term, notice, usps] [cost, based, max] [number, cluster, parameter, word, associated, collection, typically, triangle] [figure, training, corresponds, image, video, ieee, multiple, computer, pattern] [program, matrix, optimization, solution, problem, min, theoretical, denotes, sparse, nonzero, analysis, minimization, method, require, distributed] [range, encoding, threshold, left, encode, computed] [data, representative, point, proposed, regularization, zij, select, dij, text, well, selecting, clustering, error, performance, equal, dissimilarity, synthetic, machine, better, asymmetric, selection, euclidean, dataset, large, selected, small, trace, isolet]
Multi-scale Hyper-time Hardware Emulation of Human Motor Nervous System Based on Spiking Neurons using FPGA
C. Minos Niu, Sirish Nandyala, Won Joon Sohn, Terence Sanger


[order, probability, mathematical, communication, note, design] [current, control, evaluated, function, skeletal, based, state] [model, will, described, number, component, university, modeling, logic] [human, figure, produce, complex, child] [department, journal, sparse] [motor, muscle, emulation, fpga, neuron, platform, neural, spinal, spindle, sensory, nervous, system, physiological, time, emulated, implemented, spike, monosynaptic, clock, input, recruitment, unit, synapse, hardware, engineering, emg, neurophysiology, operate, combinational, spiking, nature, asynchronous, speed, clinical, force, movement, injury, division, southern, typical, biomedical, network, california, childhood, programmable, sparsely, femu, threshold] [real, differential, data, parallel, linear, well]
Perceptron Learning of SAT
Alex Flint, Matthew Blaschko


[sat, algorithm, polynomial, dpll, literal, sentence, proof, clause, theorem, margin, branching, set, setting, occurences, perceptron, horn, branch, computable, exists, propositional, length, case, mistake, planar, consider, minisat, subset, bounded] [search, function, decision, valuation, based, conference, heuristic, current, approach, total, return] [will, number, variable, update, node, tree, propagation, logic, generated] [feature, training, applied, figure, canonical, priority, score, simple, include, work, learn] [positive, procedure, problem, solution, size, application, graph, stochastic] [time, unit, negative, hardware, rule, cnf, boolean, activity, framework] [learning, yij, proposed, validation, space, performance, selected]
Learning Networks of Heterogeneous Influence
NAN DU, Le Song, Alex Smola, Ming Yuan


[true, algorithm, set, called, probability, note, observe] [function, average, parametric, continuous, based] [node, transmission, kernelcascade, diffusion, hazard, survival, event, will, likelihood, netinf, model, num, cascade, number, fji, netrate, exp, rayleigh, log, hidden, site, infected, structure, pair, inferred, parent, directed, social, process, occurs, pdf, modeling, stamp, hji, form, sji, networked, mix, piece] [figure, simple, capture, grouped, three] [optimization, estimated, problem, edge, convex, type, recover, analysis, underlying, method, corresponding, estimate, lasso] [network, time, negative] [kernel, better, data, learning, synthetic, real, performance, regularization, individual, rbf]
FastEx: Hash Clustering with Exponential Families
Amr Ahmed, Sujith Ravi, Shravan Narayanamurthy, Alex Smola


[probability, bound, note, set, measure, example, provide, instance, theorem, empirical] [total, standard, conference, cost, average, approach, key] [distribution, sampling, number, cluster, exponential, proposal, compute, conjugate, log, update, inference, sampler, likelihood, collapsed, dirichlet, document, draw, fastex, gibbs, model, prior, sample, parameter, mixture, maximum, latent, scalable, will, associated, discrete, multinomial, process, gaussian, predictive, normalization] [single, baseline, product, perform, achieved] [estimate, problem, method, family, random, estimation, high, general] [time, natural, increase] [data, clustering, hash, large, quality, fast, approximation, binary, inner, hashing, similarity, parallel, table]
Multiclass Learning with Simplex Coding
Youssef Mroueh, Tomaso Poggio, Lorenzo Rosasco, Jean-Jacques Slotine


[loss, simplex, multiclass, case, risk, considered, complexity, comparison, class, consider, online, margin, theorem, suitable, excess, explicit, set, consistency, square, statistical, derive, bayes, multicategory, induced, consistent, theory, paper, note, extension, remark, appendix, hinge, inequality, depends, empirical] [function, approach, expected, based, cost] [number, computational, distribution, computing] [map, considering, feature] [relaxation, convex, analysis, regularized, support, problem, journal, batch, corresponding, condition, solving, versus] [coding, rule, decoding, framework, fisher, noise, natural] [binary, svm, vector, large, error, learning, machine, regularization, space, performance, kernel, table, data, boosting, point, rbf]
Learning Label Trees for Probabilistic Modelling of Implicit Feedback
Andriy Mnih, Yee Whye Teh


[algorithm, probability, explicit, set, path, term, ranking, label, protocol, depend, best, assume] [based, approach, evaluation] [model, tree, implicit, log, node, factor, perplexity, will, probabilistic, number, language, modelling, scalable, bpr, distribution, structure, irrelevant, develop, evaluating, unselected, hierarchical, bmf, updating, typically, assigned] [training, level, learned, learn, child, trained, produce, single] [matrix, random, gradient, problem] [feedback, time, contribution, relevant, neural, making] [item, user, collaborative, learning, performance, selected, selection, proposed, small, unobserved, data, large, binary, test, lth, factorization, clustering, better, dataset, linear, movielens, rating, task, table, advantage]
Topic-Partitioned Multinetwork Embeddings
Peter Krafft, Juston Moore, Bruce Desmarais, Hanna Wallach


[communication, set, order, algorithm, variant, modularity, drawn] [sampled, conference, average] [model, email, topic, observed, nhc, associated, distribution, organizational, actor, county, exploratory, number, posterior, erosheva, represented, prior, concentration, modeling, parameter, predictive, enron, intended, ability, assignment, hanover, lsm, inferred, latent, joint, development, public, degree, social, assortativity, transitivity, generative, dirichlet, subnetworks, manager, link] [figure, pattern, baseline, visualization, discover, visualize, second, directly, three] [coherence, analysis, random, graph, relationship, corresponding, journal] [network, principled, prediction, typical] [data, euclidean, space, synthetic, performance, text, test, distance, demonstrate]
Graphical Models via Generalized Linear Models
Eunho Yang, Pradeep Ravikumar, Genevera Allen, Zhandong Liu


[class, set, statistical, probability, provide, case, consider, note, question, true, studied, measure] [key, function, parametric, approach, standard] [distribution, model, exponential, conditional, joint, exp, poisson, log, structure, gaussian, node, ising, markov, multinomial, parameter, form, discrete, number, will, undirected, university, rest] [learned, figure, canonical, work] [graphical, glm, graph, family, genomic, general, neighborhood, multivariate, random, suppose, constant, recovery, cancer, analysis, high, assumption, generalized, univariate, proposition, estimate, breast, edge, problem, glioblastoma, underlying, tumor, corresponding, technical, sparsity, successful, size] [network, negative, study] [data, linear, learning, selection, vector]
Memorability of Image Regions
Aditya Khosla, Jianxiong Xiao, Antonio Torralba, Aude Oliva


[algorithm, probability, ranking, set, deviation, volume] [standard, average, people, conference] [model, probabilistic, sample, process, described] [image, memorability, region, local, feature, representation, color, human, multiple, global, internal, automatic, forgotten, computer, work, memorable, semantic, learned, pyramid, score, isola, object, vision, spatial, figure, texture, corresponds, learn, remember, scene, saliency, ground, pattern, noisy, automatically, map, top, ieee, truth, single] [dictionary, type, high, size, method, rank, corresponding, propose, gradient, novel, low, close] [visual, memory, natural, experimental, framework, external, predict, effective, predicted] [performance, learning, individual, distance, predicting, dataset, large, linear, proposed]
Best Arm Identification: A Unified Approach to Fixed Budget and Fixed Confidence
Victor Gabillon, Mohammad Ghavamzadeh, Alessandro Lazaric


[arm, set, algorithm, best, budget, ugape, complexity, regret, setting, bound, case, probability, forecaster, lemma, smaller, prove, ugapeb, ugapec, bandit, fixed, returned, introduced, interval, appendix, write, note, called, result, stopping, considered, returning, upper, empirical, studied, fact, gap, extended, consider, corollary, pulled, notion, report] [max, exploration, return, conference, international, approach, optimal] [event, number, sample, parameter, allows, distribution] [accuracy, simple, figure, product, level, multiple] [problem, analysis, theoretical, arg, min, stochastic, high, step, objective, bernstein] [time, desired, account, phase, company, reported] [performance, selection, learning, existing, selected, machine, variance]
On the Sample Complexity of Robust PCA
Matthew Coudron, Gilad Lerman


[theorem, complexity, probability, proof, assume, version, lemma, term, empirical, measure, note, result, extended, order, energy, set, radius, denote, algorithm, main, series, combining, supported, mathematical, bound, theory, increasing, reaper, paper] [function, optimal, formulate] [sample, exp, component, will, distribution, compactly, mixture] [recall, bottom, work] [covariance, robust, convergence, inverse, rate, pca, matrix, subspace, generalized, underlying, estimation, norm, zhang, frobenius, high, directional, constant, estimator, estimate, convex, minimizer, recovery, outlier, wiley, size, problem, corresponding, minimization, random, lerman, assumption, close, sparse, simplicity] [explain, suggested, direct, nontrivial] [error, criterion, eigenvectors, small, approximation, ball]
A new metric on the manifold of kernel matrices with application to matrix geometric means
Suvrit Sra


[theorem, set, result, prove, lemma, inequality, studied, implies, consider, case, corollary, statistical, fact, denote, moahker, proof, minimum, paper, main, equality, iff, measure] [function, based, key, showing, stationary, faster, search, noting, proving] [compute, computing, triangle, computational, integral, turn] [local, global, second, vision, computer] [positive, matrix, symmetric, geometric, analysis, unique, diagonal, optimization, solution, multivariate, running, numerous, convex, application, covariance, crucial, theoretical, denoted, iteration, hessian] [negative, time, study, equation] [metric, space, riemannian, spd, point, geometry, algebra, manifold, hilbert, linear, kernel, chebbi, learning, cherian, distance, large, karcher, machine, isometrically, operator, worth, quantum]
Nonparametric Bayesian Inverse Reinforcement Learning for Multiple Reward Functions
Jaedeug Choi, Kee-Eung Kim


[algorithm, set, assume, probability, order, drawn, note, change] [reward, function, irl, behaviour, trajectory, approach, agent, reinforcement, mario, evd, optimal, policy, average, state, andrew, collecting, killing, birl, move, coin, lane, driving, standard, knowledge, expert, korea, evaluated, executing] [number, bayesian, distribution, cluster, posterior, generated, prior, process, mixture, model, nonparametric, dirichlet, assignment, sample, parameter, infer, assigned, markov, likelihood, normal, conditional] [multiple, single, second, feature, learned, previous, figure, three, local] [inverse, problem, gradient, graphical, denotes, random] [time, speed, left, computed, behavior] [data, learning, clustering, performance, task, distinct, binary, proposed]
Online Sum-Product Computation Over Trees
Mark Herbster, Stephen Pasteris, Fabio Vitale


[cover, vertex, algorithm, height, set, online, subtree, path, subtrees, rooted, respect, observe, notation, marginalization, introduce, covering, case, theorem, ink, dva, consider, setting, exists, exposure, lemma, cached, sequence, denote, bounded, mergeable, call, order, massimiliano, iff, notion] [optimal, atomic, belief, black, recursively, formed] [tree, hierarchical, will, update, construction, component, marginal, marginals, split, compute, propagation, computing, structure, process, number, initialization] [figure, leaf, single, root, previous, three] [decomposition, step, splitting, edge, size, application, graph, method, unique, matrix, interior] [time, boundary, potential, represent, computation, prediction] [linear, learning, data, task, selected, machine]
Sparse Approximate Manifolds for Differential Geometric MCMC
Ben Calderhead, Matyas Sustik


[statistical, consider, algorithm, order, probability, note, equivalent, circadian] [current, approximate, based, stationary, expected, deterministic, standard, employed, approach, state] [mcmc, distribution, sampling, proposal, model, structure, regularisation, parameter, markov, computational, chain, number, analytically, mala, langevin, mmala, sample, likelihood, monte, carlo, will, full, allows, deterministically, diffusion] [target, local, higher] [covariance, inverse, sparse, geometric, convergence, density, method, matrix, estimation, journal, solution, sparsity, graphical, correlation, estimating, min, globally] [adaptive, fisher, time, nonlinear, natural, position, speed] [metric, manifold, riemannian, tensor, geometry, point, tangent, differential, locally, space, approximation, vector, table, inner, challenging, scheme]
Convergence Rate Analysis of MAP Coordinate Minimization Algorithms
Ofer Meshi, Tommi Jaakkola, Amir Globerson


[algorithm, set, feasible, consider, result, theorem, denote, note, proof, fact, upper, choose, subset, energy, lemma, case, bound, analyze] [optimal, approach, function, star, max, programming, approximate, choice, belief, focus, cost] [update, will, log, integral, number, marginals, message, inference, exp, typically] [map, simple, work, ieee, local, improvement, figure, directly, compare, correspond, pattern] [primal, dual, coordinate, minimization, block, convergence, greedy, gradient, solution, stochastic, rate, analysis, objective, descent, problem, step, fista, smoothed, optimum, constant, lipschitz, accelerated, method, optimization, guaranteed, smooth, lagrangian, converge, fractional, passing, convex, decomposition] [smoothing] [linear, machine, scheme, better]
"Natural Images, Gaussian Mixtures and Dead Leaves"
Daniel Zoran, Yair Weiss


[note, set, comparison, interesting] [add, good, sampled, current, flat, stationary] [model, mixture, component, likelihood, log, structure, gaussian, number, process, modeling, generative, sampling, form, supplementary, sample, will, generated] [dead, gmm, figure, image, patch, mini, gmms, texture, gsm, ica, learned, contrast, capture, denoising, occlusion, trained, simple, side, depicts, compared, compare, variation, mask, outperforms, surprising, vision, computer, unseen, experiment, object, karklin, success, scale] [covariance, edge, pca, scalar, eigenvalue, eigenvector, matrix, random, sparse, corresponding, independent, analysis] [natural, material, neural, visual, presented, coding] [eigenvectors, performance, test, better, fourier, linear, text, small, data, learning]
Dual-Space Analysis of the Sparse Linear Model
David Wipf, Yi Wu


[case, consider, empirical, theorem, equivalent, term, exactly, minimum, note, relevance] [function, cost, optimal, max, provided, standard] [log, likelihood, will, bayesian, prior, parameter, gaussian, model, computing, update, highly, latent, factor, cluster, alternative] [view, global, local, ieee, figure, source] [type, sparse, min, arg, sparsity, dictionary, estimation, reweighted, problem, concave, solution, solving, estimate, penalty, generalized, nonzero, method, maximally, optimization, clustered, rvm, involves, iterative, general, analysis, convex, minimization, norm, satisfy, solve] [signal, lead, neural, noise, natural, principled] [learning, linear, vector, data, naturally, performance, original, small, address, well, point]
A dynamic excitatory-inhibitory network in a VLSI chip for spiking information reregistration
Juan Huo


[] [] [] [] [] [] []
ImageNet Classification with Deep Convolutional Neural Networks
Alex Krizhevsky, Ilya Sutskever, Geoff Hinton


[] [] [] [] [] [] []
Recognizing Activities by Attribute Dynamics
Weixin LI, Nuno Vasconcelos


[sequence, best, probability, note, measure, set, algorithm, generalization] [action, state, observation, sampled, expected, atomic, based] [modeling, topic, model, log, structure, gaussian, component, latent, distribution, bernoulli, hidden] [attribute, semantic, video, bof, recognition, temporal, score, representation, figure, holistic, human, learned, object, olympic, complex, popular, kds, weizmann, computer, feature, accuracy, ykt, image, occurrence, learn, inspired, performed] [pca, analysis, matrix, subspace, method] [dynamic, activity, time, lds, visual, natural, system, noise, motion, computed, reported] [binary, kernel, learning, performance, similarity, table, space, vector, proposed, dataset, linear, data]
Compressive Sensing MRI with Wavelet Tree Sparsity
Chen Chen, Junzhou Huang


[algorithm, set, statistical, result, partial, complexity, best] [total, conference, cost, international, based, good, average] [tree, structure, sampling, model, number, log, prior, bayesian] [wavelet, image, snr, ieee, figure, imaging, structured, variation, slep, mri, previous, cpu, magnetic, compare, fcsa, computer, bottom, resonance, medical, recpf, tvcmri] [reconstruction, sparsity, problem, sparse, solve, compressive, convex, method, sensing, amp, general, convergence, arg, solution, solved, compressed, gradient, conducted, subproblems, slow, numerous, scanning, inverse, fista, reconstructed, denotes, iterative, achieves, robust, simpler] [group, signal, time, brain, visual, left, processing] [proposed, real, original, fast, performance, existing, fourier, better, learning]
Bayesian Nonparametric Modeling of Suicide Attempts
Francisco Ruiz, Isabel Valera, Carlos Blanco, Fernando Perez-Cruz


[probability, risk, active, question, survey, set, consider, element, order] [observation, black, standard, people, function] [suicide, latent, model, distribution, ibp, attempt, number, gibbs, discrete, hidden, gaussian, log, posterior, znk, prevention, sampler, alcohol, nesarc, process, xnd, allows, prior, inference, university, indian, topic, likelihood, dirichlet, buffet, modeling, laplace, answered, base, nonparametric, nth, american, attempting, sampling, prevalence] [feature, map, improve, medical, figure, pixel, weighting, carlos, applied, madrid] [matrix, denoted, journal, low, propose, iteration, random] [population, processing, white, integrate] [data, table, vector, database, binary, approximation, national, well]
Stochastic optimization and sparse statistical recovery: Optimal algorithms for high dimensions
Alekh Agarwal, Sahand Negahban, Martin Wainwright


[algorithm, theorem, set, loss, bound, statistical, probability, version, convexity, example, length, setting, sequence, paper, supported, corollary, design, main, subset, logistic, observe, assume, mirror] [optimal, function, initial, approach, focus, expected, choice] [log, form, update, sample, parameter, will] [epoch, work, exploit] [stochastic, convergence, objective, optimization, rate, sparse, gradient, method, regularized, convex, problem, strong, radar, optimum, averaging, dual, constant, sparsity, condition, assumption, min, descent, eda, high, lipschitz, scaling, random, meaning, journal, solution] [natural] [error, learning, regularization, machine, vector, common, large, data, linear]
Latent Coincidence Analysis: A Hidden Variable Model for Distance Metric Learning
Matthew Der, Lawrence Saul


[class, set, algorithm, example, note, label, comparison, length, denote] [approach, explore, simplest, conference, international] [latent, model, posterior, gaussian, variable, distribution, factor, parameter, observed, mixture, inference, update, likelihood, bayesian, compute, maximum, prior, lda, odds] [training, learn, target, simple, transformation, scale, level, figure, larger, train, compare] [pca, analysis, high, multivariate, dimensional, matrix] [computed, noise, principal, rule] [lca, labeled, data, metric, learning, distance, linear, dimensionality, knn, error, coincidence, test, space, lmnn, bbc, mnist, isolet, variance, neighbor, table, machine, large, projected, closer, reduction, binary, similarity, nearest, multiway]
Hierarchical Optimistic Region Selection driven by Curiosity
Odalric-Ambrym Maillard


[def, algorithm, order, regret, allocation, consider, case, term, optimistic, bound, setting, loss, oudeyer, notion, main, carpentier, assume, active, provide, curiosity, baranes, introduce, theorem, bubeck, antos, proof, paper, probability, upper, arm, interesting, choose, result, note, quantity, trivial, exists, spirit] [max, unknown, continuous, function, reinforcement, assumed, goal, optimal, standard, future, state, argmax] [partition, sample, sampling, hierarchical, node, tree, process, split] [region, driven, corresponds, leaf, work] [problem, assumption, estimation, analysis, objective, lipschitz, min, step, constant, positive, corresponding] [cell, time, difference, autonomous, intrinsic] [learning, space, variance, performance, metric, dimension, selection, error, well, point, select, challenging, better, binary]
Finite Sample Convergence Rates of Zero-Order Stochastic Optimization Methods
John Duchi, Michael Jordan, Martin Wainwright, Andre Wibisono


[lower, mirror, minimax, set, online, theorem, respect, bound, algorithm, corollary, bandit, lemma, proof, quantity, attained, radius, universal, sequence, appendix, choose, exists, denote, implies, assume, uniform, sharp] [function, based, optimal, additional, choice, introduction, continuous, expectation] [log, distribution, parameter, full, separation, sample, factor] [apply, accuracy, work, second] [convergence, optimization, gradient, stochastic, convex, random, descent, assumption, norm, rate, constant, method, proposition, analysis, estimator, agarwal, proximal, problem, subgradient, achieve, lipschitz, objective, nesterov, dependence] [study] [learning, vector, selection, dimension, small, unbiased, machine, logarithmic, better, error, annual]
Trajectory-Based Short-Sighted Probabilistic Planning
Felipe Trevizan, Manuela Veloso


[probability, set, path, algorithm, sequence, consider, applicable, asymptotically, theorem, example] [ssipp, state, planning, ssp, goal, policy, optimal, uncertainty, ssps, search, action, planner, tireworld, manage, international, based, uct, initial, car, bellman, heuristic, conference, reached, approach, cost, tire, reach, trajectory, execution, admissible, shortest, spare, reachable, reaching, mdpsim, terminates, ptimal, round, pmax, reachability, scheduling, move, olver, determinization, intelligence] [probabilistic, tr