title
string
paper_url
string
authors
list
type
string
primary_area
string
abstract
large_string
keywords
list
TL;DR
large_string
submission_number
int64
arxiv_id
string
arxiv_id_source
string
embedding
list
No Oops, You Won’t Do It Again: Mechanisms for Self-correction in Crowdsourcing
https://proceedings.mlr.press/v48/shaha16.html
[ "Nihar Shah", "Dengyong Zhou" ]
null
null
Crowdsourcing is a very popular means of obtaining the large amounts of labeled data that modern machine learning methods require. Although cheap and fast to obtain, crowdsourced labels suffer from significant amounts of error, thereby degrading the performance of downstream machine learning tasks. With the goal of imp...
[]
null
1
null
null
[ 0.016049640253186226, -0.046185098588466644, -0.0412549190223217, 0.05934743955731392, 0.03168085590004921, 0.028628017753362656, -0.0007948160637170076, 0.0032226755283772945, -0.01950726844370365, -0.02377437613904476, -0.025955351069569588, 0.007462760899215937, -0.05249916389584541, -0...
Stochastically Transitive Models for Pairwise Comparisons: Statistical and Computational Issues
https://proceedings.mlr.press/v48/shahb16.html
[ "Nihar Shah", "Sivaraman Balakrishnan", "Aditya Guntuboyina", "Martin Wainwright" ]
null
null
There are various parametric models for analyzing pairwise comparison data, including the Bradley-Terry-Luce (BTL) and Thurstone models, but their reliance on strong parametric assumptions is limiting. In this work, we study a flexible model for pairwise comparisons, under which the probabilities of outcomes are requir...
[]
null
2
1510.05610
title_snapshot
[ 0.004539427347481251, 0.00046155709424056113, -0.03101767972111702, 0.03098016418516636, 0.00941398274153471, 0.029763078317046165, 0.04651281237602234, 0.020628521218895912, -0.029632313176989555, -0.04220075532793999, -0.00630434462800622, 0.008519772440195084, -0.049723070114851, -0.008...
Uprooting and Rerooting Graphical Models
https://proceedings.mlr.press/v48/weller16.html
[ "Adrian Weller" ]
null
null
We show how any binary pairwise model may be “uprooted” to a fully symmetric model, wherein original singleton potentials are transformed to potentials on edges to an added variable, and then “rerooted” to a new model on the original number of variables. The new model is essentially equivalent to the original model, wi...
[]
null
3
null
null
[ -0.02048604190349579, -0.014286518096923828, -0.01206506323069334, 0.04948331043124199, 0.04566033557057381, 0.029070397838950157, 0.039931125938892365, 0.003164127469062805, -0.019136546179652214, -0.03751114383339882, -0.002573162317276001, -0.015580400824546814, -0.10433162748813629, -0...
A Deep Learning Approach to Unsupervised Ensemble Learning
https://proceedings.mlr.press/v48/shaham16.html
[ "Uri Shaham", "Xiuyuan Cheng", "Omer Dror", "Ariel Jaffe", "Boaz Nadler", "Joseph Chang", "Yuval Kluger" ]
null
null
We show how deep learning methods can be applied in the context of crowdsourcing and unsupervised ensemble learning. First, we prove that the popular model of Dawid and Skene, which assumes that all classifiers are conditionally independent, is \em equivalent to a Restricted Boltzmann Machine (RBM) with a single hidden...
[]
null
4
1602.02285
title_snapshot
[ -0.0021537907887250185, -0.0330783911049366, -0.026398908346891403, 0.017305657267570496, 0.02859531342983246, 0.003819872858002782, 0.013161986134946346, -0.01623387262225151, -0.007286138366907835, -0.036315303295850754, -0.009735709987580776, 0.019838374108076096, -0.07304416596889496, ...
Revisiting Semi-Supervised Learning with Graph Embeddings
https://proceedings.mlr.press/v48/yanga16.html
[ "Zhilin Yang", "William Cohen", "Ruslan Salakhudinov" ]
null
null
We present a semi-supervised learning framework based on graph embeddings. Given a graph between instances, we train an embedding for each instance to jointly predict the class label and the neighborhood context in the graph. We develop both transductive and inductive variants of our method. In the transductive variant...
[]
null
5
1603.08861
title_snapshot
[ 0.019278256222605705, -0.05720868334174156, -0.0007009539403952658, 0.06489668041467667, 0.02668454311788082, -0.005920158699154854, 0.022233963012695312, -0.009703190065920353, 0.013904727064073086, -0.010938158258795738, -0.02462773397564888, 0.013512915000319481, -0.06942135095596313, 0...
Guided Cost Learning: Deep Inverse Optimal Control via Policy Optimization
https://proceedings.mlr.press/v48/finn16.html
[ "Chelsea Finn", "Sergey Levine", "Pieter Abbeel" ]
null
null
Reinforcement learning can acquire complex behaviors from high-level specifications. However, defining a cost function that can be optimized effectively and encodes the correct task is challenging in practice. We explore how inverse optimal control (IOC) can be used to learn behaviors from demonstrations, with applicat...
[]
null
6
1603.00448
title_snapshot
[ -0.055754199624061584, -0.02245897240936756, -0.017620133236050606, 0.03808717802166939, 0.05979994684457779, 0.023990573361516, 0.013370188884437084, -0.006632007192820311, -0.01436575036495924, -0.029624801129102707, -0.006726985797286034, 0.039138682186603546, -0.041411224752664566, -0....
Diversity-Promoting Bayesian Learning of Latent Variable Models
https://proceedings.mlr.press/v48/xiea16.html
[ "Pengtao Xie", "Jun Zhu", "Eric Xing" ]
null
null
In learning latent variable models (LVMs), it is important to effectively capture infrequent patterns and shrink model size without sacrificing modeling power. Various studies have been done to “diversify” a LVM, which aim to learn a diverse set of latent components in LVMs. Most existing studies fall into a frequentis...
[]
null
7
1711.08770
title_snapshot
[ 0.024088185280561447, -0.02726484276354313, -0.01855311542749405, 0.037542443722486496, 0.039236441254615784, 0.04616463929414749, 0.03907802328467369, -0.032109662890434265, -0.05012373998761177, -0.05015086010098457, -0.005311583634465933, 0.027902042493224144, -0.05682188645005226, 0.01...
Additive Approximations in High Dimensional Nonparametric Regression via the SALSA
https://proceedings.mlr.press/v48/kandasamy16.html
[ "Kirthevasan Kandasamy", "Yaoliang Yu" ]
null
null
High dimensional nonparametric regression is an inherently difficult problem with known lower bounds depending exponentially in dimension. A popular strategy to alleviate this curse of dimensionality has been to use additive models of \emphfirst order, which model the regression function as a sum of independent functio...
[]
null
8
1602.00287
title_snapshot
[ -0.04459894821047783, -0.02900131233036518, -0.0043005612678825855, 0.02317877486348152, 0.03428392484784126, 0.05513444170355797, 0.03168627619743347, -0.03641004487872124, -0.048737846314907074, -0.01733480766415596, -0.014579985290765762, 0.01665448024868965, -0.06511152535676956, 0.020...
Hawkes Processes with Stochastic Excitations
https://proceedings.mlr.press/v48/leea16.html
[ "Young Lee", "Kar Wai Lim", "Cheng Soon Ong" ]
null
null
We propose an extension to Hawkes processes by treating the levels of self-excitation as a stochastic differential equation. Our new point process allows better approximation in application domains where events and intensities accelerate each other with correlated levels of contagion. We generalize a recent algorithm f...
[]
null
9
1609.06831
title_snapshot
[ 0.023488003760576248, 0.010872096754610538, -0.02567795105278492, 0.01697307825088501, 0.01276223175227642, 0.044524554163217545, 0.01840949058532715, 0.008853939361870289, -0.013240472413599491, -0.0717816948890686, 0.03148649260401726, -0.008345426060259342, -0.03616659343242645, 0.01881...
Data-driven Rank Breaking for Efficient Rank Aggregation
https://proceedings.mlr.press/v48/khetan16.html
[ "Ashish Khetan", "Sewoong Oh" ]
null
null
Rank aggregation systems collect ordinal preferences from individuals to produce a global ranking that represents the social preference. To reduce the computational complexity of learning the global ranking, a common practice is to use rank-breaking. Individuals’ preferences are broken into pairwise comparisons and the...
[]
null
10
1601.05495
title_snapshot
[ -0.03346613049507141, -0.021080298349261284, 0.03389883413910866, 0.03032802790403366, -0.0035526708234101534, 0.00571184977889061, 0.03251805528998375, -0.020725863054394722, -0.026314275339245796, -0.0370183102786541, -0.012822197750210762, 0.0075957803055644035, -0.0931367501616478, -0....
Dropout distillation
https://proceedings.mlr.press/v48/bulo16.html
[ "Samuel Rota Bulò", "Lorenzo Porzi", "Peter Kontschieder" ]
null
null
Dropout is a popular stochastic regularization technique for deep neural networks that works by randomly dropping (i.e. zeroing) units from the network during training. This randomization process allows to implicitly train an ensemble of exponentially many networks sharing the same parametrization, which should be aver...
[]
null
11
null
null
[ 0.02979990839958191, -0.021482940763235092, -0.030659519135951996, 0.04813311621546745, 0.03167062997817993, 0.005472642835229635, 0.03178110718727112, -0.0028838792350143194, -0.031734809279441833, -0.03615821897983551, -0.023267114534974098, -0.03393431380391121, -0.04781473055481911, -0...
Metadata-conscious anonymous messaging
https://proceedings.mlr.press/v48/fanti16.html
[ "Giulia Fanti", "Peter Kairouz", "Sewoong Oh", "Kannan Ramchandran", "Pramod Viswanath" ]
null
null
Anonymous messaging platforms like Whisper and Yik Yak allow users to spread messages over a network (e.g., a social network) without revealing message authorship to other users. The spread of messages on these platforms can be modeled by a diffusion process over a graph. Recent advances in network analysis have reveal...
[]
null
12
null
null
[ 0.0024781820829957724, -0.015057757496833801, -0.0016038973117247224, 0.06293334811925888, 0.05535956099629402, -0.01905258744955063, 0.056549716740846634, -0.00309097021818161, -0.022968990728259087, -0.02299983985722065, 0.022845564410090446, -0.017914189025759697, -0.04907386004924774, ...
The Teaching Dimension of Linear Learners
https://proceedings.mlr.press/v48/liua16.html
[ "Ji Liu", "Xiaojin Zhu", "Hrag Ohannessian" ]
null
null
Teaching dimension is a learning theoretic quantity that specifies the minimum training set size to teach a target model to a learner. Previous studies on teaching dimension focused on version-space learners which maintain all hypotheses consistent with the training data, and cannot be applied to modern machine learner...
[]
null
13
1512.02181
title_snapshot
[ -0.021388724446296692, -0.02014179900288582, -0.013430776074528694, 0.0054953028447926044, 0.052125245332717896, 0.045924112200737, 0.05402737483382225, -0.025676289573311806, -0.05542416498064995, -0.016853362321853638, -0.023071013391017914, 0.009654135443270206, -0.05895253270864487, 0....
Truthful Univariate Estimators
https://proceedings.mlr.press/v48/caragiannis16.html
[ "Ioannis Caragiannis", "Ariel Procaccia", "Nisarg Shah" ]
null
null
We revisit the classic problem of estimating the population mean of an unknown single-dimensional distribution from samples, taking a game-theoretic viewpoint. In our setting, samples are supplied by strategic agents, who wish to pull the estimate as close as possible to their own value. In this setting, the sample mea...
[]
null
14
null
null
[ -0.043953657150268555, -0.021461937576532364, 0.0012732894392684102, 0.024543922394514084, 0.044759660959243774, 0.058254584670066833, 0.04039255529642105, 0.015406455844640732, -0.04474661499261856, -0.05094024911522865, 0.009431521408259869, -0.0018484449246898293, -0.0773414745926857, -...
Why Regularized Auto-Encoders learn Sparse Representation?
https://proceedings.mlr.press/v48/arpita16.html
[ "Devansh Arpit", "Yingbo Zhou", "Hung Ngo", "Venu Govindaraju" ]
null
null
Sparse distributed representation is the key to learning useful features in deep learning algorithms, because not only it is an efficient mode of data representation, but also – more importantly – it captures the generation process of most real world data. While a number of regularized auto-encoders (AE) enforce sparsi...
[]
null
15
1505.05561
title_snapshot
[ 0.01339083444327116, -0.033473871648311615, -0.00650459760800004, 0.04260958358645439, 0.034400615841150284, 0.04304890334606171, 0.030421370640397072, -0.0056999498046934605, -0.04583554342389107, -0.04766470193862915, 0.012575777247548103, -0.025083286687731743, -0.04621428996324539, 0.0...
k-variates++: more pluses in the k-means++
https://proceedings.mlr.press/v48/nock16.html
[ "Richard Nock", "Raphael Canyasse", "Roksana Boreli", "Frank Nielsen" ]
null
null
k-means++ seeding has become a de facto standard for hard clustering algorithms. In this paper, our first contribution is a two-way generalisation of this seeding, k-variates++, that includes the sampling of general densities rather than just a discrete set of Dirac densities anchored at the point locations, *and* a ge...
[]
null
16
1602.01198
title_snapshot
[ 0.0007390952087007463, -0.010761240497231483, 0.0466497503221035, 0.05277000367641449, 0.028000278398394585, 0.042584583163261414, 0.03605087846517563, -0.027528807520866394, -0.035591430962085724, -0.03000016137957573, -0.01848318986594677, -0.04512818157672882, -0.05170688405632973, -0.0...
Multi-Player Bandits – a Musical Chairs Approach
https://proceedings.mlr.press/v48/rosenski16.html
[ "Jonathan Rosenski", "Ohad Shamir", "Liran Szlak" ]
null
null
We consider a variant of the stochastic multi-armed bandit problem, where multiple players simultaneously choose from the same set of arms and may collide, receiving no reward. This setting has been motivated by problems arising in cognitive radio networks, and is especially challenging under the realistic assumption t...
[]
null
17
1512.02866
title_snapshot
[ -0.036026693880558014, -0.02248143032193184, 0.004528840072453022, 0.04140562564134598, 0.03185545653104782, 0.008417193777859211, 0.02154688537120819, 0.01761302724480629, -0.05103247985243797, -0.07083766162395477, -0.03309241682291031, 0.024771833792328835, -0.057384878396987915, -0.016...
The Information Sieve
https://proceedings.mlr.press/v48/steeg16.html
[ "Greg Ver Steeg", "Aram Galstyan" ]
null
null
We introduce a new framework for unsupervised learning of representations based on a novel hierarchical decomposition of information. Intuitively, data is passed through a series of progressively fine-grained sieves. Each layer of the sieve recovers a single latent factor that is maximally informative about multivariat...
[]
null
18
1507.02284
title_snapshot
[ -0.005985788069665432, -0.0007923291414044797, -0.013388621620833874, 0.0351240299642086, 0.04614221677184105, 0.03741772100329399, 0.010253495536744595, -0.002342741470783949, -0.03255349397659302, -0.03842499852180481, 0.0056196595542132854, 0.01816626451909542, -0.06392953544855118, 0.0...
Deep Speech 2 : End-to-End Speech Recognition in English and Mandarin
https://proceedings.mlr.press/v48/amodei16.html
[ "Dario Amodei", "Sundaram Ananthanarayanan", "Rishita Anubhai", "Jingliang Bai", "Eric Battenberg", "Carl Case", "Jared Casper", "Bryan Catanzaro", "Qiang Cheng", "Guoliang Chen", "Jie Chen", "Jingdong Chen", "Zhijie Chen", "Mike Chrzanowski", "Adam Coates", "Greg Diamos", "Ke Ding",...
null
null
We show that an end-to-end deep learning approach can be used to recognize either English or Mandarin Chinese speech–two vastly different languages. Because it replaces entire pipelines of hand-engineered components with neural networks, end-to-end learning allows us to handle a diverse variety of speech including nois...
[]
null
19
1512.02595
title_snapshot
[ -0.010732780210673809, -0.000028952345019206405, -0.050893303006887436, 0.0240120030939579, 0.019484324380755424, 0.04763880744576454, 0.030544545501470566, 0.04513523727655411, 0.007925833575427532, -0.04455889016389847, -0.016065193340182304, 0.010548066347837448, -0.040731046348810196, ...
On the Consistency of Feature Selection With Lasso for Non-linear Targets
https://proceedings.mlr.press/v48/zhanga16.html
[ "Yue Zhang", "Weihong Guo", "Soumya Ray" ]
null
null
An important question in feature selection is whether a selection strategy recovers the “true” set of features, given enough data. We study this question in the context of the popular Least Absolute Shrinkage and Selection Operator (Lasso) feature selection strategy. In particular, we consider the scenario when the mod...
[]
null
20
null
null
[ -0.04101887717843056, 0.002380233258008957, 0.022217130288481712, 0.002145796548575163, 0.05868290364742279, 0.04993313550949097, 0.048750560730695724, -0.0060010007582604885, -0.04916304722428322, -0.06200537830591202, -0.018977981060743332, 0.022947506979107857, -0.06449344754219055, -0....
Minimum Regret Search for Single- and Multi-Task Optimization
https://proceedings.mlr.press/v48/metzen16.html
[ "Jan Hendrik Metzen" ]
null
null
We propose minimum regret search (MRS), a novel acquisition function for Bayesian optimization. MRS bears similarities with information-theoretic approaches such as entropy search (ES). However, while ES aims in each query at maximizing the information gain with respect to the global maximum, MRS aims at minimizing the...
[]
null
21
1602.01064
title_snapshot
[ -0.029181616380810738, 0.009532676078379154, -0.011512970551848412, 0.05551661550998688, 0.043813955038785934, 0.05288880318403244, 0.020049842074513435, -0.005475063342601061, -0.028006449341773987, -0.06966013461351395, -0.023937538266181946, 0.03792516514658928, -0.03961917385458946, -0...
CryptoNets: Applying Neural Networks to Encrypted Data with High Throughput and Accuracy
https://proceedings.mlr.press/v48/gilad-bachrach16.html
[ "Ran Gilad-Bachrach", "Nathan Dowlin", "Kim Laine", "Kristin Lauter", "Michael Naehrig", "John Wernsing" ]
null
null
Applying machine learning to a problem which involves medical, financial, or other types of sensitive data, not only requires accurate predictions but also careful attention to maintaining data privacy and security. Legal and ethical requirements may prevent the use of cloud-based machine learning solutions for such ta...
[]
null
22
null
null
[ 0.011980206705629826, -0.0229856688529253, -0.02364874817430973, 0.05881819128990173, 0.06613654643297195, 0.04793817177414894, 0.0037496425211429596, -0.005400415975600481, -0.03557632490992546, -0.02968488819897175, 0.001362815615721047, -0.01522368285804987, -0.029594529420137405, 0.013...
The Variational Nystrom method for large-scale spectral problems
https://proceedings.mlr.press/v48/vladymyrov16.html
[ "Max Vladymyrov", "Miguel Carreira-Perpinan" ]
null
null
Spectral methods for dimensionality reduction and clustering require solving an eigenproblem defined by a sparse affinity matrix. When this matrix is large, one seeks an approximate solution. The standard way to do this is the Nystrom method, which first solves a small eigenproblem considering only a subset of landmark...
[]
null
23
null
null
[ -0.027580585330724716, 0.0067268600687384605, 0.022195493802428246, 0.01665448397397995, 0.04772873595356941, 0.0301330778747797, 0.019594090059399605, -0.014783242717385292, -0.02363223396241665, -0.047612641006708145, -0.01651330478489399, 0.005019000731408596, -0.08785294741392136, 0.01...
Multi-Bias Non-linear Activation in Deep Neural Networks
https://proceedings.mlr.press/v48/lia16.html
[ "Hongyang Li", "Wanli Ouyang", "Xiaogang Wang" ]
null
null
As a widely used non-linear activation, Rectified Linear Unit (ReLU) separates noise and signal in a feature map by learning a threshold or bias. However, we argue that the classification of noise and signal not only depends on the magnitude of responses, but also the context of how the feature responses would be used ...
[]
null
24
1604.00676
title_snapshot
[ 0.004826795309782028, -0.0035249164793640375, 0.011193991638720036, 0.0267245601862669, 0.010803031735122204, 0.05371465906500816, 0.013426772318780422, 0.0033605671487748623, -0.030491605401039124, -0.053779009729623795, 0.011492514982819557, -0.0017459867522120476, -0.07601454854011536, ...
Asymmetric Multi-task Learning Based on Task Relatedness and Loss
https://proceedings.mlr.press/v48/leeb16.html
[ "Giwoong Lee", "Eunho Yang", "Sung Hwang" ]
null
null
We propose a novel multi-task learning method that can minimize the effect of negative transfer by allowing asymmetric transfer between the tasks based on task relatedness as well as the amount of individual task losses, which we refer to as Asymmetric Multi-task Learning (AMTL). To tackle this problem, we couple multi...
[]
null
25
null
null
[ 0.0259024016559124, -0.023925254121422768, -0.01173397060483694, 0.0019078553887084126, 0.04670156165957451, 0.011751865968108177, 0.02574990689754486, -0.007959622889757156, -0.01525367796421051, -0.04775865748524666, -0.007987412624061108, 0.03488198667764664, -0.06745085120201111, -0.03...
Accurate Robust and Efficient Error Estimation for Decision Trees
https://proceedings.mlr.press/v48/fan16.html
[ "Lixin Fan" ]
null
null
This paper illustrates a novel approach to the estimation of generalization error of decision tree classifiers. We set out the study of decision tree errors in the context of consistency analysis theory, which proved that the Bayes error can be achieved only if when the number of data samples thrown into each leaf node...
[]
null
26
null
null
[ -0.0010551807936280966, 0.01497105322778225, -0.02659706212580204, 0.027383198961615562, 0.04663956165313721, 0.03857114911079407, 0.053586509078741074, -0.021155420690774918, -0.010161359794437885, -0.01863075979053974, -0.015566457062959671, 0.010603757575154305, -0.08583786338567734, -0...
Fast Stochastic Algorithms for SVD and PCA: Convergence Properties and Convexity
https://proceedings.mlr.press/v48/shamira16.html
[ "Ohad Shamir" ]
null
null
We study the convergence properties of the VR-PCA algorithm introduced by (Shamir, 2015) for fast computation of leading singular vectors. We prove several new results, including a formal analysis of a block version of the algorithm, and convergence from random initialization. We also make a few observations of indepen...
[]
null
27
1507.08788
title_snapshot
[ -0.00766275729984045, -0.018243543803691864, 0.04150408133864403, 0.03279722481966019, 0.005946473218500614, 0.047124456614255905, 0.0328739732503891, 0.02374100685119629, -0.04101092740893364, -0.05430169776082039, -0.021326303482055664, -0.02615320309996605, -0.057569265365600586, 0.0011...
Convergence of Stochastic Gradient Descent for PCA
https://proceedings.mlr.press/v48/shamirb16.html
[ "Ohad Shamir" ]
null
null
We consider the problem of principal component analysis (PCA) in a streaming stochastic setting, where our goal is to find a direction of approximate maximal variance, based on a stream of i.i.d. data points in R^d. A simple and computationally cheap algorithm for this is stochastic gradient descent (SGD), which increm...
[]
null
28
1509.09002
title_snapshot
[ -0.013155601918697357, -0.028140557929873466, 0.0285887960344553, 0.04777344688773155, 0.020501045510172844, 0.03464381769299507, 0.03077765554189682, 0.017368730157613754, -0.028790997341275215, -0.04698358103632927, -0.013114545494318008, -0.03616032004356384, -0.08909936249256134, -0.00...
Dealbreaker: A Nonlinear Latent Variable Model for Educational Data
https://proceedings.mlr.press/v48/lan16.html
[ "Andrew Lan", "Tom Goldstein", "Richard Baraniuk", "Christoph Studer" ]
null
null
Statistical models of student responses on assessment questions, such as those in homeworks and exams, enable educators and computer-based personalized learning systems to gain insights into students’ knowledge using machine learning. Popular student-response models, including the Rasch model and item response theory m...
[]
null
29
null
null
[ 0.001156630809418857, -0.03150830790400505, -0.037989504635334015, 0.0569998137652874, 0.0428733266890049, 0.023304227739572525, 0.012018309906125069, 0.005904870107769966, -0.03570907562971115, -0.010966245085000992, -0.0030307427514344454, 0.024696381762623787, -0.026743872091174126, 0.0...
A Kernelized Stein Discrepancy for Goodness-of-fit Tests
https://proceedings.mlr.press/v48/liub16.html
[ "Qiang Liu", "Jason Lee", "Michael Jordan" ]
null
null
We derive a new discrepancy statistic for measuring differences between two probability distributions based on combining Stein’s identity and the reproducing kernel Hilbert space theory. We apply our result to test how well a probabilistic model fits a set of observations, and derive a new class of powerful goodness-of...
[]
null
30
1602.03253
title_judge
[ -0.0359625518321991, -0.002952774753794074, 0.0017792652361094952, 0.03838757798075676, 0.07095402479171753, 0.03595879673957825, 0.028824234381318092, -0.01337865274399519, -0.0064848302863538265, -0.04624965786933899, 0.0070426808670163155, 0.007241752929985523, -0.06434408575296402, -0....
Variable Elimination in the Fourier Domain
https://proceedings.mlr.press/v48/xue16.html
[ "Yexiang Xue", "Stefano Ermon", "Ronan Le Bras", "Carla", "Bart Selman" ]
null
null
The ability to represent complex high dimensional probability distributions in a compact form is one of the key insights in the field of graphical models. Factored representations are ubiquitous in machine learning and lead to major computational advantages. We explore a different type of compact representation based o...
[]
null
31
1508.04032
title_snapshot
[ -0.0282637570053339, 0.019467810168862343, 0.004193165339529514, 0.03992262855172157, 0.05432684347033501, 0.03854277357459068, 0.031035035848617554, -0.019772231578826904, -0.021485144272446632, -0.03113454580307007, -0.004363853484392166, 0.016844192519783974, -0.06908382475376129, 0.025...
Low-Rank Matrix Approximation with Stability
https://proceedings.mlr.press/v48/lib16.html
[ "Dongsheng Li", "Chao Chen", "Qin Lv", "Junchi Yan", "Li Shang", "Stephen Chu" ]
null
null
Low-rank matrix approximation has been widely adopted in machine learning applications with sparse data, such as recommender systems. However, the sparsity of the data, incomplete and noisy, introduces challenges to the algorithm stability – small changes in the training data may significantly change the models. As a r...
[]
null
32
null
null
[ -0.009844282642006874, -0.027317708358168602, 0.013421754352748394, 0.02120836265385151, 0.048833463340997696, 0.01732194982469082, 0.012754784896969795, -0.031367454677820206, -0.03437300771474838, -0.04985127970576286, -0.007467674557119608, 0.004010976757854223, -0.07369086891412735, 0....
Linking losses for density ratio and class-probability estimation
https://proceedings.mlr.press/v48/menon16.html
[ "Aditya Menon", "Cheng Soon Ong" ]
null
null
Given samples from two densities p and q, density ratio estimation (DRE) is the problem of estimating the ratio p/q. Two popular discriminative approaches to DRE are KL importance estimation (KLIEP), and least squares importance fitting (LSIF). In this paper, we show that KLIEP and LSIF both employ class-probability es...
[]
null
33
null
null
[ 0.011193804442882538, 0.014153732918202877, -0.011462829075753689, 0.03311936557292938, 0.02973291091620922, 0.056692712008953094, -0.00013133998436387628, -0.020716136321425438, -0.010960434563457966, -0.043511077761650085, -0.02080991305410862, 0.018907776102423668, -0.06192690134048462, ...
Stochastic Variance Reduction for Nonconvex Optimization
https://proceedings.mlr.press/v48/reddi16.html
[ "Sashank J. Reddi", "Ahmed Hefny", "Suvrit Sra", "Barnabas Poczos", "Alex Smola" ]
null
null
We study nonconvex finite-sum problems and analyze stochastic variance reduced gradient (SVRG) methods for them. SVRG and related methods have recently surged into prominence for convex optimization given their edge over stochastic gradient descent (SGD); but their theoretical analysis almost exclusively assumes convex...
[]
null
34
1603.06160
title_snapshot
[ -0.015393907204270363, -0.025638416409492493, 0.014847414568066597, 0.03502030298113823, 0.009916620329022408, 0.04626649618148804, 0.036967650055885315, 0.014190878719091415, -0.02542358636856079, -0.05718196555972099, -0.00881313905119896, -0.02016083337366581, -0.05623304843902588, -0.0...
Hierarchical Variational Models
https://proceedings.mlr.press/v48/ranganath16.html
[ "Rajesh Ranganath", "Dustin Tran", "David Blei" ]
null
null
Black box variational inference allows researchers to easily prototype and evaluate an array of models. Recent advances allow such algorithms to scale to high dimensions. However, a central question remains: How to specify an expressive variational distribution that maintains efficient computation? To address this, we ...
[]
null
35
1511.02386
title_snapshot
[ -0.016300301998853683, 0.027014344930648804, -0.0037904568016529083, 0.04136084020137787, 0.02413281239569187, 0.053330693393945694, 0.023189421743154526, 0.002309247152879834, -0.04596107453107834, -0.04381530359387398, -0.008016424253582954, -0.00337175908498466, -0.05246935039758682, 0....
Hierarchical Span-Based Conditional Random Fields for Labeling and Segmenting Events in Wearable Sensor Data Streams
https://proceedings.mlr.press/v48/adams16.html
[ "Roy Adams", "Nazir Saleheen", "Edison Thomaz", "Abhinav Parate", "Santosh Kumar", "Benjamin Marlin" ]
null
null
The field of mobile health (mHealth) has the potential to yield new insights into health and behavior through the analysis of continuously recorded data from wearable health and activity sensors. In this paper, we present a hierarchical span-based conditional random field model for the key problem of jointly detecting ...
[]
null
36
null
null
[ 0.020219314843416214, -0.016346532851457596, 0.0026333287823945284, 0.03786498308181763, 0.04116488993167877, 0.02522461861371994, 0.016830746084451675, -0.008203290402889252, -0.03885943815112114, -0.029237806797027588, -0.019327502697706223, 0.016132675111293793, -0.045834146440029144, -...
Binary embeddings with structured hashed projections
https://proceedings.mlr.press/v48/choromanska16.html
[ "Anna Choromanska", "Krzysztof Choromanski", "Mariusz Bojarski", "Tony Jebara", "Sanjiv Kumar", "Yann LeCun" ]
null
null
We consider the hashing mechanism for constructing binary embeddings, that involves pseudo-random projections followed by nonlinear (sign function) mappings. The pseudo-random projection is described by a matrix, where not all entries are independent random variables but instead a fixed “budget of randomness” is distri...
[]
null
37
1511.05212
title_snapshot
[ 0.01485168281942606, -0.014332226477563381, 0.004180802032351494, 0.06084088236093521, 0.03257116675376892, 0.030897799879312515, 0.019969940185546875, -0.013558543287217617, -0.017668915912508965, -0.0209804680198431, -0.030458567664027214, -0.01730533502995968, -0.05762076377868652, -0.0...
A Variational Analysis of Stochastic Gradient Algorithms
https://proceedings.mlr.press/v48/mandt16.html
[ "Stephan Mandt", "Matthew Hoffman", "David Blei" ]
null
null
Stochastic Gradient Descent (SGD) is an important algorithm in machine learning. With constant learning rates, it is a stochastic process that, after an initial phase of convergence, generates samples from a stationary distribution. We show that SGD with constant rates can be effectively used as an approximate posterio...
[]
null
38
1602.02666
title_snapshot
[ -0.02962823584675789, 0.008381940424442291, 0.0030124441254884005, 0.03362521156668663, 0.04697873815894127, 0.035151366144418716, 0.03334817290306091, 0.01374107412993908, -0.01985863596200943, -0.041578713804483414, 0.004243700765073299, -0.00591413164511323, -0.04811122640967369, 0.0078...
Adaptive Sampling for SGD by Exploiting Side Information
https://proceedings.mlr.press/v48/gopal16.html
[ "Siddharth Gopal" ]
null
null
This paper proposes a new mechanism for sampling training instances for stochastic gradient descent (SGD) methods by exploiting any side-information associated with the instances (for e.g. class-labels) to improve convergence. Previous methods have either relied on sampling from a distribution defined over training ins...
[]
null
39
null
null
[ -0.05229562893509865, -0.03767071291804314, 0.0008240173920057714, 0.04620717465877533, 0.043830420821905136, 0.04307461902499199, 0.03413872420787811, -0.00849725492298603, -0.010693523101508617, -0.03990297392010689, -0.0079742930829525, -0.019295819103717804, -0.07015885412693024, -0.01...
Learning from Multiway Data: Simple and Efficient Tensor Regression
https://proceedings.mlr.press/v48/yu16.html
[ "Rose Yu", "Yan Liu" ]
null
null
Tensor regression has shown to be advantageous in learning tasks with multi-directional relatedness. Given massive multiway data, traditional methods are often too slow to operate on or suffer from memory bottleneck. In this paper, we introduce subsampled tensor projected gradient to solve the problem. Our algorithm is...
[]
null
40
1607.02535
title_snapshot
[ -0.0028925109654664993, -0.035211749374866486, 0.011780480854213238, 0.01124864537268877, 0.036888208240270615, 0.02888207882642746, 0.0133920693770051, -0.01317133754491806, -0.020285077393054962, -0.061804890632629395, 0.0018930223304778337, 0.014214283786714077, -0.06527364999055862, 0....
A Distributed Variational Inference Framework for Unifying Parallel Sparse Gaussian Process Regression Models
https://proceedings.mlr.press/v48/hoang16.html
[ "Trong Nghia Hoang", "Quang Minh Hoang", "Bryan Kian Hsiang Low" ]
null
null
This paper presents a novel distributed variational inference framework that unifies many parallel sparse Gaussian process regression (SGPR) models for scalable hyperparameter learning with big data. To achieve this, our framework exploits a structure of correlated noise process model that represents the observation no...
[]
null
41
null
null
[ -0.005969093646854162, -0.0073219710029661655, -0.0009253026801161468, 0.029550010338425636, 0.027884898707270622, 0.05980971083045006, 0.037210576236248016, 0.0037480201572179794, -0.0253719724714756, -0.03232802078127861, 0.004255489446222782, 0.004524230491369963, -0.07619484513998032, ...
Online Stochastic Linear Optimization under One-bit Feedback
https://proceedings.mlr.press/v48/zhangb16.html
[ "Lijun Zhang", "Tianbao Yang", "Rong Jin", "Yichi Xiao", "Zhi-hua Zhou" ]
null
null
In this paper, we study a special bandit setting of online stochastic linear optimization, where only one-bit of information is revealed to the learner at each round. This problem has found many applications including online advertisement and online recommendation. We assume the binary feedback is a random variable gen...
[]
null
42
1509.07728
title_snapshot
[ 0.0060350652784109116, -0.0037365108728408813, 0.021804355084896088, 0.030774660408496857, 0.03924137353897095, 0.0615893192589283, 0.026280255988240242, 0.04515005648136139, -0.00692368671298027, -0.045265376567840576, -0.029670728370547295, -0.03724910318851471, -0.07020395994186401, -0....
Adaptive Algorithms for Online Convex Optimization with Long-term Constraints
https://proceedings.mlr.press/v48/jenatton16.html
[ "Rodolphe Jenatton", "Jim Huang", "Cedric Archambeau" ]
null
null
We present an adaptive online gradient descent algorithm to solve online convex optimization problems with long-term constraints, which are constraints that need to be satisfied when accumulated over a finite number of rounds T, but can be violated in intermediate rounds. For some user-defined trade-off parameter βin (...
[]
null
43
1512.07422
title_snapshot
[ -0.042298946529626846, -0.02230759523808956, 0.0020926385186612606, 0.020831704139709473, 0.0405067503452301, 0.056965962052345276, 0.01479338388890028, 0.024538952857255936, -0.029420262202620506, -0.024158397689461708, -0.04109509661793709, -0.0025768980849534273, -0.05139753222465515, -...
Actively Learning Hemimetrics with Applications to Eliciting User Preferences
https://proceedings.mlr.press/v48/singla16.html
[ "Adish Singla", "Sebastian Tschiatschek", "Andreas Krause" ]
null
null
Motivated by an application of eliciting users’ preferences, we investigate the problem of learning hemimetrics, i.e., pairwise distances among a set of n items that satisfy triangle inequalities and non-negativity constraints. In our application, the (asymmetric) distances quantify private costs a user incurs when sub...
[]
null
44
1605.07144
title_snapshot
[ -0.00531178992241621, 0.0038600796833634377, 0.030010275542736053, -0.003356017405167222, 0.024030692875385284, 0.016537368297576904, -0.018730713054537773, -0.017539503052830696, -0.01750093139708042, -0.04628307744860649, -0.020046565681695938, 0.0233856663107872, -0.06512099504470825, 0...
Learning Simple Algorithms from Examples
https://proceedings.mlr.press/v48/zaremba16.html
[ "Wojciech Zaremba", "Tomas Mikolov", "Armand Joulin", "Rob Fergus" ]
null
null
We present an approach for learning simple algorithms such as copying, multi-digit addition and single digit multiplication directly from examples. Our framework consists of a set of interfaces, accessed by a controller. Typical interfaces are 1-D tapes or 2-D grids that hold the input and output data. For the controll...
[]
null
45
1511.07275
title_snapshot
[ -0.029302386566996574, -0.006519613321870565, -0.035175420343875885, 0.04691507667303085, 0.04301628842949867, 0.02062021568417549, 0.004577199928462505, 0.010430943220853806, -0.03181947022676468, -0.016443992033600807, 0.022074034437537193, 0.0035361982882022858, -0.07430563122034073, -0...
Learning Physical Intuition of Block Towers by Example
https://proceedings.mlr.press/v48/lerer16.html
[ "Adam Lerer", "Sam Gross", "Rob Fergus" ]
null
null
Wooden blocks are a common toy for infants, allowing them to develop motor skills and gain intuition about the physical behavior of the world. In this paper, we explore the ability of deep feed-forward models to learn such intuitive physics. Using a 3D game engine, we create small towers of wooden blocks whose stabilit...
[]
null
46
1603.01312
title_snapshot
[ 0.01143726147711277, -0.012730157002806664, -0.005954164080321789, 0.02601774036884308, 0.03201071918010712, 0.017370279878377914, 0.020283734425902367, 0.009688650257885456, -0.04115649685263634, -0.03473367914557457, -0.03278341144323349, -0.0505782812833786, -0.0646541565656662, -0.0084...
Structure Learning of Partitioned Markov Networks
https://proceedings.mlr.press/v48/liuc16.html
[ "Song Liu", "Taiji Suzuki", "Masashi Sugiyama", "Kenji Fukumizu" ]
null
null
We learn the structure of a Markov Network between two groups of random variables from joint observations. Since modelling and learning the full MN structure may be hard, learning the links between two groups directly may be a preferable option. We introduce a novel concept called the \emphpartitioned ratio whose facto...
[]
null
47
1504.00624
title_snapshot
[ -0.015841949731111526, -0.017001986503601074, -0.035452213138341904, 0.034281279891729355, 0.034516841173172, 0.03381986543536186, 0.03654180467128754, -0.0026052361354231834, -0.029656153172254562, -0.036002643406391144, 0.04275623336434364, -0.016479957848787308, -0.07644445449113846, -0...
Tracking Slowly Moving Clairvoyant: Optimal Dynamic Regret of Online Learning with True and Noisy Gradient
https://proceedings.mlr.press/v48/yangb16.html
[ "Tianbao Yang", "Lijun Zhang", "Rong Jin", "Jinfeng Yi" ]
null
null
This work focuses on dynamic regret of online convex optimization that compares the performance of online learning to a clairvoyant who knows the sequence of loss functions in advance and hence selects the minimizer of the loss function at each step. By assuming that the clairvoyant moves slowly (i.e., the minimizers c...
[]
null
48
1605.04638
title_snapshot
[ -0.013978558592498302, -0.01937071792781353, 0.009590852074325085, 0.04019622877240181, 0.03989465907216072, 0.031191714107990265, 0.03950073942542076, 0.013981172814965248, -0.020290151238441467, -0.06688977777957916, -0.023493485525250435, 0.008281845599412918, -0.04766609147191048, -0.0...
Beyond CCA: Moment Matching for Multi-View Models
https://proceedings.mlr.press/v48/podosinnikova16.html
[ "Anastasia Podosinnikova", "Francis Bach", "Simon Lacoste-Julien" ]
null
null
We introduce three novel semi-parametric extensions of probabilistic canonical correlation analysis with identifiability guarantees. We consider moment matching techniques for estimation in these models. For that, by drawing explicit links between the new models and a discrete version of independent component analysis ...
[]
null
49
1602.09013
title_snapshot
[ 0.02488267794251442, 0.0016796194249764085, -0.008333312347531319, 0.009077680297195911, 0.027783837169408798, 0.04681416228413582, 0.0378638319671154, 0.021469153463840485, -0.003031684085726738, -0.05420701950788498, -0.007539114914834499, -0.001695247134193778, -0.07956670969724655, -0....
Fast methods for estimating the Numerical rank of large matrices
https://proceedings.mlr.press/v48/ubaru16.html
[ "Shashanka Ubaru", "Yousef Saad" ]
null
null
We present two computationally inexpensive techniques for estimating the numerical rank of a matrix, combining powerful tools from computational linear algebra. These techniques exploit three key ingredients. The first is to approximate the projector on the non-null invariant subspace of the matrix by using a polynomia...
[]
null
50
null
null
[ -0.03551621735095978, -0.00472305528819561, 0.011257925070822239, 0.009337397292256355, 0.014283315278589725, -0.009642960503697395, 0.009547427296638489, -0.02912834845483303, -0.05249572917819023, -0.05425069108605385, -0.01066602859646082, -0.00023447364219464362, -0.0521782748401165, 0...
Unsupervised Deep Embedding for Clustering Analysis
https://proceedings.mlr.press/v48/xieb16.html
[ "Junyuan Xie", "Ross Girshick", "Ali Farhadi" ]
null
null
Clustering is central to many data-driven application domains and has been studied extensively in terms of distance functions and grouping algorithms. Relatively little work has focused on learning representations for clustering. In this paper, we propose Deep Embedded Clustering (DEC), a method that simultaneously lea...
[]
null
51
1511.06335
title_snapshot
[ -0.016961360350251198, -0.04597700759768486, -0.018969595432281494, 0.03700363636016846, 0.06946951150894165, 0.048329584300518036, 0.007248608395457268, -0.0100529994815588, -0.0010417640442028642, -0.027091478928923607, -0.026561148464679718, 0.003571063280105591, -0.03012201003730297, 0...
Efficient Private Empirical Risk Minimization for High-dimensional Learning
https://proceedings.mlr.press/v48/kasiviswanathan16.html
[ "Shiva Prasad Kasiviswanathan", "Hongxia Jin" ]
null
null
Dimensionality reduction is a popular approach for dealing with high dimensional data that leads to substantial computational savings. Random projections are a simple and effective method for universal dimensionality reduction with rigorous theoretical guarantees. In this paper, we theoretically study the problem of di...
[]
null
52
null
null
[ -0.024662956595420837, 0.00849885307252407, 0.01724383234977722, 0.056660979986190796, 0.03228534013032913, 0.03965688496828079, 0.026356011629104614, -0.06329318135976791, -0.027141107246279716, -0.042489193379879, -0.023834846913814545, -0.0350584015250206, -0.05790811777114868, 0.014948...
Parameter Estimation for Generalized Thurstone Choice Models
https://proceedings.mlr.press/v48/vojnovic16.html
[ "Milan Vojnovic", "Seyoung Yun" ]
null
null
We consider the maximum likelihood parameter estimation problem for a generalized Thurstone choice model, where choices are from comparison sets of two or more items. We provide tight characterizations of the mean square error, as well as necessary and sufficient conditions for correct classification when each item bel...
[]
null
53
null
null
[ -0.016325101256370544, -0.007230299524962902, -0.035056523978710175, 0.021155914291739464, 0.022253673523664474, 0.04618929326534271, 0.013769206590950489, 0.045356106013059616, -0.032663505524396896, -0.046385787427425385, 0.00036500729038380086, 0.00706322630867362, -0.0616675540804863, ...
Large-Margin Softmax Loss for Convolutional Neural Networks
https://proceedings.mlr.press/v48/liud16.html
[ "Weiyang Liu", "Yandong Wen", "Zhiding Yu", "Meng Yang" ]
null
null
Cross-entropy loss together with softmax is arguably one of the most common used supervision components in convolutional neural networks (CNNs). Despite its simplicity, popularity and excellent performance, the component does not explicitly encourage discriminative learning of features. In this paper, we propose a gene...
[]
null
54
1612.02295
title_snapshot
[ -0.01135939359664917, -0.02970796637237072, 0.03232910856604576, 0.004368640016764402, 0.022330744192004204, 0.00002628928268677555, -0.010343444533646107, 0.00918224360793829, -0.01108529418706894, -0.03227442130446434, -0.01237962394952774, -0.0037110131233930588, -0.07337818294763565, 0...
A Random Matrix Approach to Echo-State Neural Networks
https://proceedings.mlr.press/v48/couillet16.html
[ "Romain Couillet", "Gilles Wainrib", "Hafiz Tiomoko Ali", "Harry Sevi" ]
null
null
Recurrent neural networks, especially in their linear version, have provided many qualitative insights on their performance under different configurations. This article provides, through a novel random matrix framework, the quantitative counterpart of these performance results, specifically in the case of echo-state ne...
[]
null
55
null
null
[ -0.021878905594348907, -0.030094198882579803, -0.029742777347564697, 0.01794680766761303, 0.036935605108737946, 0.05047832429409027, 0.032417140901088715, 0.03147801011800766, -0.050444431602954865, -0.05090980976819992, 0.018885420635342598, -0.017719825729727745, -0.06215308606624603, -0...
Supervised and Semi-Supervised Text Categorization using LSTM for Region Embeddings
https://proceedings.mlr.press/v48/johnson16.html
[ "Rie Johnson", "Tong Zhang" ]
null
null
One-hot CNN (convolutional neural network) has been shown to be effective for text categorization (Johnson & Zhang, 2015). We view it as a special case of a general framework which jointly trains a linear model with a non-linear feature generator consisting of ‘text region embedding + pooling’. Under this framework, we...
[]
null
56
1602.02373
title_snapshot
[ 0.0004232731880620122, -0.06674784421920776, -0.024946918711066246, 0.06407590210437775, 0.05154558643698692, 0.008367080241441727, 0.015151341445744038, 0.021358300000429153, -0.015813598409295082, -0.018957138061523438, -0.026757875457406044, 0.007392777130007744, -0.050627026706933975, ...
Optimality of Belief Propagation for Crowdsourced Classification
https://proceedings.mlr.press/v48/ok16.html
[ "Jungseul Ok", "Sewoong Oh", "Jinwoo Shin", "Yung Yi" ]
null
null
Crowdsourcing systems are popular for solving large-scale labelling tasks with low-paid (or even non-paid) workers. We study the problem of recovering the true labels from noisy crowdsourced labels under the popular Dawid-Skene model. To address this inference problem, several algorithms have recently been proposed, bu...
[]
null
57
null
null
[ -0.0030993346590548754, -0.019461438059806824, -0.00672710919752717, 0.05109753459692001, 0.02484891563653946, 0.02443922869861126, 0.014352771453559399, -0.002829499775543809, -0.031249692663550377, -0.017881829291582108, -0.01769215613603592, 0.010643670335412025, -0.07625709474086761, -...
Stability of Controllers for Gaussian Process Forward Models
https://proceedings.mlr.press/v48/vinogradska16.html
[ "Julia Vinogradska", "Bastian Bischoff", "Duy Nguyen-Tuong", "Anne Romer", "Henner Schmidt", "Jan Peters" ]
null
null
Learning control has become an appealing alternative to the derivation of control laws based on classic control theory. However, a major shortcoming of learning control is the lack of performance guarantees which prevents its application in many real-world scenarios. As a step in this direction, we provide a stability ...
[]
null
58
null
null
[ -0.04598592594265938, 0.002582112094387412, -0.009182716719806194, 0.02106316201388836, 0.05833371728658676, 0.01376634556800127, 0.008287055417895317, 0.01427615899592638, -0.013261592946946621, -0.05338532477617264, -0.004586400464177132, 0.021183926612138748, -0.08146482706069946, -0.00...
Learning privately from multiparty data
https://proceedings.mlr.press/v48/hamm16.html
[ "Jihun Hamm", "Yingjun Cao", "Mikhail Belkin" ]
null
null
Learning a classifier from private data distributed across multiple parties is an important problem that has many potential applications. How can we build an accurate and differentially private global classifier by combining locally-trained classifiers from different parties, without access to any party’s private data?...
[]
null
59
1602.03552
title_snapshot
[ 0.01622636243700981, -0.011451033875346184, -0.012602690607309341, 0.06194935366511345, 0.02827712707221508, 0.0016007546801120043, 0.03538589924573898, -0.028972743079066277, -0.016174819320440292, -0.018175406381487846, 0.009962661191821098, -0.0063713314011693, -0.07662094384431839, 0.0...
Network Morphism
https://proceedings.mlr.press/v48/wei16.html
[ "Tao Wei", "Changhu Wang", "Yong Rui", "Chang Wen Chen" ]
null
null
We present a systematic study on how to morph a well-trained neural network to a new one so that its network function can be completely preserved. We define this as network morphism in this research. After morphing a parent network, the child network is expected to inherit the knowledge from its parent network and also...
[]
null
60
1603.01670
title_snapshot
[ -0.012061630375683308, -0.03260437771677971, 0.00284911529161036, 0.042061708867549896, 0.048496268689632416, 0.04994312301278114, 0.006671107839792967, -0.008136607706546783, -0.035098545253276825, -0.06806408613920212, 0.009358174167573452, -0.014817874878644943, -0.030817091464996338, 0...
A Kronecker-factored approximate Fisher matrix for convolution layers
https://proceedings.mlr.press/v48/grosse16.html
[ "Roger Grosse", "James Martens" ]
null
null
Second-order optimization methods such as natural gradient descent have the potential to speed up training of neural networks by correcting for the curvature of the loss function. Unfortunately, the exact natural gradient is impractical to compute for large models, and most approximations either require an expensive it...
[]
null
61
1602.01407
title_snapshot
[ -0.00292141642421484, -0.07024259120225906, 0.007933329790830612, 0.05203501880168915, 0.03820575028657913, 0.04727794602513313, 0.012277581728994846, 0.009636634960770607, -0.015874594449996948, -0.05129536986351013, 0.017805567011237144, 0.0007658859249204397, -0.035383161157369614, 0.00...
Experimental Design on a Budget for Sparse Linear Models and Applications
https://proceedings.mlr.press/v48/ravi16.html
[ "Sathya Narayanan Ravi", "Vamsi Ithapu", "Sterling Johnson", "Vikas Singh" ]
null
null
Budget constrained optimal design of experiments is a classical problem in statistics. Although the optimal design literature is very mature, few efficient strategies are available when these design problems appear in the context of sparse linear models commonly encountered in high dimensional machine learning and stat...
[]
null
62
null
null
[ -0.013463671319186687, 0.008272168226540089, -0.030115297064185143, 0.02776811085641384, 0.04633050411939621, 0.03161177411675453, 0.034417133778333664, -0.0047760228626430035, -0.03735481947660446, -0.05147501081228256, 0.0033843780402094126, -0.0009915768168866634, -0.0695386752486229, 0...
Minding the Gaps for Block Frank-Wolfe Optimization of Structured SVMs
https://proceedings.mlr.press/v48/osokin16.html
[ "Anton Osokin", "Jean-Baptiste Alayrac", "Isabella Lukasewitz", "Puneet Dokania", "Simon Lacoste-Julien" ]
null
null
In this paper, we propose several improvements on the block-coordinate Frank-Wolfe (BCFW) algorithm from Lacoste-Julien et al. (2013) recently used to optimize the structured support vector machine (SSVM) objective in the context of structured prediction, though it has wider applications. The key intuition behind our i...
[]
null
63
1605.09346
title_snapshot
[ 0.01005295105278492, -0.026181630790233612, 0.025981906801462173, 0.024589605629444122, 0.04633856192231178, 0.019820023328065872, 0.046103984117507935, -0.027012165635824203, -0.015113113448023796, -0.039350591599941254, -0.007106272038072348, 0.007010511122643948, -0.06598495692014694, -...
Exact Exponent in Optimal Rates for Crowdsourcing
https://proceedings.mlr.press/v48/gaoa16.html
[ "Chao Gao", "Yu Lu", "Dengyong Zhou" ]
null
null
Crowdsourcing has become a popular tool for labeling large datasets. This paper studies the optimal error rate for aggregating crowdsourced labels provided by a collection of amateur workers. Under the Dawid-Skene probabilistic model, we establish matching upper and lower bounds with an exact exponent mI(\pi), where m ...
[]
null
64
1605.07696
title_snapshot
[ -0.000012056224477419164, 0.0025549258571118116, -0.0066739353351294994, 0.020454267039895058, 0.00638550566509366, 0.0065368483774363995, 0.02261732704937458, 0.012676368467509747, -0.04779540002346039, -0.016729578375816345, -0.003923856187611818, -0.03736993297934532, -0.08783602714538574...
Augmenting Supervised Neural Networks with Unsupervised Objectives for Large-scale Image Classification
https://proceedings.mlr.press/v48/zhangc16.html
[ "Yuting Zhang", "Kibok Lee", "Honglak Lee" ]
null
null
Unsupervised learning and supervised learning are key research topics in deep learning. However, as high-capacity supervised neural networks trained with a large amount of labels have achieved remarkable success in many computer vision tasks, the availability of large-scale labeled images reduced the significance of un...
[]
null
65
1606.06582
title_snapshot
[ 0.026586182415485382, -0.04328549653291702, -0.02236069366335869, 0.02896287851035595, 0.02587253972887993, 0.005695111118257046, 0.023050915449857712, 0.0005454565398395061, -0.026372967287898064, -0.04327739030122757, -0.015521397814154625, 0.00012495815462898463, -0.0695762187242508, 0....
Online Low-Rank Subspace Clustering by Basis Dictionary Pursuit
https://proceedings.mlr.press/v48/shen16.html
[ "Jie Shen", "Ping Li", "Huan Xu" ]
null
null
Low-Rank Representation (LRR) has been a significant method for segmenting data that are generated from a union of subspaces. It is also known that solving LRR is challenging in terms of time complexity and memory footprint, in that the size of the nuclear norm regularized matrix is n-by-n (where n is the number of sam...
[]
null
66
1503.08356
title_judge
[ -0.025017738342285156, -0.021137289702892303, 0.017390964552760124, 0.019646408036351204, 0.03474759683012962, 0.038447070866823196, 0.014085437171161175, -0.0010644667781889439, -0.0436469167470932, -0.03009958565235138, -0.029855867847800255, -0.01253938302397728, -0.05582130327820778, -...
A Self-Correcting Variable-Metric Algorithm for Stochastic Optimization
https://proceedings.mlr.press/v48/curtis16.html
[ "Frank Curtis" ]
null
null
An algorithm for stochastic (convex or nonconvex) optimization is presented. The algorithm is variable-metric in the sense that, in each iteration, the step is computed through the product of a symmetric positive definite scaling matrix and a stochastic (mini-batch) gradient of the objective function, where the sequenc...
[]
null
67
null
null
[ -0.012280414812266827, 0.003317109774798155, 0.02027171663939953, 0.00468608271330595, 0.04354410991072655, 0.0868058130145073, 0.03832075372338295, 0.019709507003426552, -0.046912677586078644, -0.040567681193351746, -0.016933942213654518, -0.006677624303847551, -0.036967795342206955, -0.0...
Stochastic Quasi-Newton Langevin Monte Carlo
https://proceedings.mlr.press/v48/simsekli16.html
[ "Umut Simsekli", "Roland Badeau", "Taylan Cemgil", "Gaël Richard" ]
null
null
Recently, Stochastic Gradient Markov Chain Monte Carlo (SG-MCMC) methods have been proposed for scaling up Monte Carlo computations to large data problems. Whilst these approaches have proven useful in many applications, vanilla SG-MCMC might suffer from poor mixing rates when random variables exhibit strong couplings ...
[]
null
68
1602.03442
title_snapshot
[ -0.025145450606942177, 0.0029751122929155827, 0.01715618371963501, 0.04168475419282913, 0.03157372772693634, 0.040008656680583954, 0.029288874939084053, 0.0010525828693062067, -0.040664687752723694, -0.06994334608316422, 0.026985086500644684, -0.007793304976075888, -0.05084763467311859, -0...
Doubly Robust Off-policy Value Evaluation for Reinforcement Learning
https://proceedings.mlr.press/v48/jiang16.html
[ "Nan Jiang", "Lihong Li" ]
null
null
We study the problem of off-policy value evaluation in reinforcement learning (RL), where one aims to estimate the value of a new policy based on data collected by a different policy. This problem is often a critical step when applying RL to real-world problems. Despite its importance, existing general methods either h...
[]
null
69
1511.03722
title_snapshot
[ -0.018879368901252747, -0.00650375708937645, -0.009310666471719742, 0.03760305419564247, 0.03864974528551102, 0.019800055772066116, 0.016465183347463608, -0.004170150030404329, -0.01505262590944767, -0.028196265920996666, -0.021710367873311043, 0.015464081428945065, -0.09245750308036804, -...
Fast Rate Analysis of Some Stochastic Optimization Algorithms
https://proceedings.mlr.press/v48/qua16.html
[ "Chao Qu", "Huan Xu", "Chong Ong" ]
null
null
In this paper, we revisit three fundamental and popular stochastic optimization algorithms (namely, Online Proximal Gradient, Regularized Dual Averaging method and ADMM with online proximal gradient) and analyze their convergence speed under conditions weaker than those in literature. In particular, previous works show...
[]
null
70
null
null
[ -0.03353961184620857, -0.017305457964539528, 0.015316275879740715, 0.011490292847156525, 0.03620695322751999, 0.03992851823568344, 0.01601722277700901, 0.030212754383683205, -0.016089649870991707, -0.032192472368478775, -0.008161305449903011, -0.013116329908370972, -0.05463553965091705, -0...
Fast k-Nearest Neighbour Search via Dynamic Continuous Indexing
https://proceedings.mlr.press/v48/lic16.html
[ "Ke Li", "Jitendra Malik" ]
null
null
Existing methods for retrieving k-nearest neighbours suffer from the curse of dimensionality. We argue this is caused in part by inherent deficiencies of space partitioning, which is the underlying strategy used by most existing methods. We devise a new strategy that avoids partitioning the vector space and present a n...
[]
null
71
1512.00442
title_snapshot
[ -0.04539147764444351, -0.03954307362437248, -0.0013792160898447037, 0.02253063954412937, 0.03305814787745476, 0.06336697190999985, 0.017061036080121994, -0.014494994655251503, -0.017357828095555305, -0.04798320680856705, -0.01370320376008749, -0.04564521461725235, -0.051651619374752045, 0....
Smooth Imitation Learning for Online Sequence Prediction
https://proceedings.mlr.press/v48/le16.html
[ "Hoang Le", "Andrew Kang", "Yisong Yue", "Peter Carr" ]
null
null
We study the problem of smooth imitation learning for online sequence prediction, where the goal is to train a policy that can smoothly imitate demonstrated behavior in a dynamic and continuous environment in response to online, sequential context input. Since the mapping from context to behavior is often complex, we t...
[]
null
72
1606.00968
title_snapshot
[ 0.002603545319288969, -0.02222442999482155, 0.006987273693084717, 0.038958899676799774, 0.05398216471076012, 0.04381704330444336, 0.04559039697051048, 0.018925899639725685, -0.03041306510567665, -0.009674531407654285, -0.0163540318608284, 0.005429988261312246, -0.0703752413392067, -0.01014...
Community Recovery in Graphs with Locality
https://proceedings.mlr.press/v48/chena16.html
[ "Yuxin Chen", "Govinda Kamath", "Changho Suh", "David Tse" ]
null
null
Motivated by applications in domains such as social networks and computational biology, we study the problem of community recovery in graphs with locality. In this problem, pairwise noisy measurements of whether two nodes are in the same community or different communities come mainly or exclusively from nearby nodes ra...
[]
null
73
1602.03828
title_snapshot
[ -0.00441339984536171, -0.01349885668605566, 0.0015802140114828944, 0.05499187111854553, 0.039708249270915985, 0.020864617079496384, 0.041278768330812454, 0.007078846450895071, -0.01716497354209423, -0.029210926964879036, 0.019276121631264687, -0.04796789586544037, -0.07453948259353638, 0.0...
Variance Reduction for Faster Non-Convex Optimization
https://proceedings.mlr.press/v48/allen-zhua16.html
[ "Zeyuan Allen-Zhu", "Elad Hazan" ]
null
null
We consider the fundamental problem in non-convex optimization of efficiently reaching a stationary point. In contrast to the convex case, in the long history of this basic problem, the only known theoretical results on first-order non-convex optimization remain to be full gradient descent that converges in O(1/\vareps...
[]
null
74
1603.05643
title_snapshot
[ -0.02603430673480034, -0.013611793518066406, 0.022740015760064125, 0.03283500298857689, 0.027052730321884155, 0.053386349231004715, 0.019534273073077202, 0.0003643473901320249, -0.030113518238067627, -0.04621882364153862, -0.005105090327560902, 0.0004863899666815996, -0.05165933817625046, ...
Loss factorization, weakly supervised learning and label noise robustness
https://proceedings.mlr.press/v48/patrini16.html
[ "Giorgio Patrini", "Frank Nielsen", "Richard Nock", "Marcello Carioni" ]
null
null
We prove that the empirical risk of most well-known loss functions factors into a linear term aggregating all labels with a term that is label free, and can further be expressed by sums of the same loss. This holds true even for non-smooth, non-convex losses and in any RKHS. The first term is a (kernel) mean operator —...
[]
null
75
1602.02450
title_snapshot
[ -0.01736685261130333, -0.020772049203515053, 0.005022473633289337, 0.037275657057762146, 0.024167194962501526, 0.019389266148209572, 0.020385924726724625, -0.01070288848131895, 0.0033355685882270336, -0.029272980988025665, -0.0197889506816864, 0.022035561501979828, -0.07361806929111481, 0....
Analysis of Deep Neural Networks with Extended Data Jacobian Matrix
https://proceedings.mlr.press/v48/wanga16.html
[ "Shengjie Wang", "Abdel-rahman Mohamed", "Rich Caruana", "Jeff Bilmes", "Matthai Plilipose", "Matthew Richardson", "Krzysztof Geras", "Gregor Urban", "Ozlem Aslan" ]
null
null
Deep neural networks have achieved great successes on various machine learning tasks, however, there are many open fundamental questions to be answered. In this paper, we tackle the problem of quantifying the quality of learned wights of different networks with possibly different architectures, going beyond considering...
[]
null
76
null
null
[ -0.03428785875439644, -0.04672189801931381, -0.011637978255748749, 0.034511249512434006, 0.030953384935855865, 0.030160533264279366, 0.03106839209794998, -0.016385797411203384, -0.0400584451854229, -0.07198220491409302, -0.005922235082834959, 0.004398568999022245, -0.04495522379875183, 0.0...
Doubly Decomposing Nonparametric Tensor Regression
https://proceedings.mlr.press/v48/imaizumi16.html
[ "Masaaki Imaizumi", "Kohei Hayashi" ]
null
null
Nonparametric extension of tensor regression is proposed. Nonlinearity in a high-dimensional tensor space is broken into simple local functions by incorporating low-rank tensor decomposition. Compared to naive nonparametric approaches, our formulation considerably improves the convergence rate of estimation while maint...
[]
null
77
1506.05967
title_snapshot
[ -0.04284115508198738, -0.006654670462012291, 0.004134907852858305, 0.02183915115892887, 0.021081479266285896, 0.04923494532704353, 0.016208073124289513, -0.043064918369054794, -0.007891133427619934, -0.04062632471323013, -0.0034871778916567564, 0.04816678166389465, -0.07759138196706772, 0....
Hyperparameter optimization with approximate gradient
https://proceedings.mlr.press/v48/pedregosa16.html
[ "Fabian Pedregosa" ]
null
null
Most models in machine learning contain at least one hyperparameter to control for model complexity. Choosing an appropriate set of hyperparameters is both crucial in terms of model accuracy and computationally challenging. In this work we propose an algorithm for the optimization of continuous hyperparameters using in...
[]
null
78
1602.02355
title_snapshot
[ -0.03787749633193016, -0.006609741132706404, 0.011773325502872467, 0.04157821089029312, 0.032675422728061676, 0.04324861615896225, 0.037857115268707275, -0.036814264953136444, -0.02545013651251793, -0.021811535581946373, -0.02686479315161705, 0.0018592558335512877, -0.03992117941379547, 0....
SDCA without Duality, Regularization, and Individual Convexity
https://proceedings.mlr.press/v48/shalev-shwartza16.html
[ "Shai Shalev-Shwartz" ]
null
null
Stochastic Dual Coordinate Ascent is a popular method for solving regularized loss minimization for the case of convex losses. We describe variants of SDCA that do not require explicit regularization and do not rely on duality. We prove linear convergence rates even if individual loss functions are non-convex, as long ...
[]
null
79
1602.01582
title_snapshot
[ -0.014749855734407902, -0.005359547678381205, 0.000847517978399992, 0.051057327538728714, 0.02016480453312397, 0.059056177735328674, -0.00008063371205935255, 0.00035295257112011313, -0.0037860118318349123, -0.06082238629460335, -0.014314794912934303, -0.019538836553692818, -0.039206031709909...
Heteroscedastic Sequences: Beyond Gaussianity
https://proceedings.mlr.press/v48/anava16.html
[ "Oren Anava", "Shie Mannor" ]
null
null
We address the problem of sequential prediction in the heteroscedastic setting, when both the signal and its variance are assumed to depend on explanatory variables. By applying regret minimization techniques, we devise an efficient online learning algorithm for the problem, without assuming that the error terms comply...
[]
null
80
null
null
[ -0.009361052885651588, 0.004651293158531189, 0.016164176166057587, 0.010914705693721771, 0.04996374994516373, 0.03496461734175682, 0.02205806039273739, 0.03163386508822441, -0.03014211170375347, -0.051101986318826675, 0.0039895083755254745, 0.025641897693276405, -0.04845018312335014, 0.012...
A Neural Autoregressive Approach to Collaborative Filtering
https://proceedings.mlr.press/v48/zheng16.html
[ "Yin Zheng", "Bangsheng Tang", "Wenkui Ding", "Hanning Zhou" ]
null
null
This paper proposes CF-NADE, a neural autoregressive architecture for collaborative filtering (CF) tasks, which is inspired by the Restricted Boltzmann Machine (RBM) based CF model and the Neural Autoregressive Distribution Estimator (NADE). We first describe the basic CF-NADE model for CF tasks. Then we propose to imp...
[]
null
81
1605.09477
title_snapshot
[ 0.029183024540543556, -0.028907490894198418, 0.029669104143977165, 0.004201640374958515, 0.04501556605100632, 0.02393311634659767, 0.013751942664384842, -0.004397122655063868, -0.0009193351725116372, -0.0699537992477417, 0.009028950706124306, 0.01884457841515541, -0.035786956548690796, -0....
On the Quality of the Initial Basin in Overspecified Neural Networks
https://proceedings.mlr.press/v48/safran16.html
[ "Itay Safran", "Ohad Shamir" ]
null
null
Deep learning, in the form of artificial neural networks, has achieved remarkable practical success in recent years, for a variety of difficult machine learning applications. However, a theoretical explanation for this remains a major open problem, since training neural networks involves optimizing a highly non-convex ...
[]
null
82
1511.04210
title_snapshot
[ -0.04030495509505272, -0.017246168106794357, 0.009937346912920475, 0.04138391837477684, 0.04808056727051735, 0.057581063359975815, -0.000033524775062687695, 0.018600227311253548, -0.03991056606173515, -0.043404776602983475, -0.0029906723648309708, -0.01475560199469328, -0.05726516246795654, ...
Primal-Dual Rates and Certificates
https://proceedings.mlr.press/v48/dunner16.html
[ "Celestine Dünner", "Simone Forte", "Martin Takac", "Martin Jaggi" ]
null
null
We propose an algorithm-independent framework to equip existing optimization methods with primal-dual certificates. Such certificates and corresponding rate of convergence guarantees are important for practitioners to diagnose progress, in particular in machine learning applications. We obtain new primal-dual convergen...
[]
null
83
1602.05205
title_snapshot
[ -0.017214976251125336, -0.004390654154121876, 0.006415117532014847, 0.03143306076526642, 0.03267086669802666, 0.035070355981588364, 0.022201135754585266, -0.0016643523704260588, -0.016792042180895805, -0.030354319140315056, -0.01881217584013939, 0.015947600826621056, -0.06079232320189476, ...
Minimizing the Maximal Loss: How and Why
https://proceedings.mlr.press/v48/shalev-shwartzb16.html
[ "Shai Shalev-Shwartz", "Yonatan Wexler" ]
null
null
A commonly used learning rule is to approximately minimize the \emphaverage loss over the training set. Other learning algorithms, such as AdaBoost and hard-SVM, aim at minimizing the \emphmaximal loss over the training set. The average loss is more popular, particularly in deep learning, due to three main reasons. Fir...
[]
null
84
1602.01690
title_snapshot
[ -0.021990224719047546, -0.03828561678528786, 0.012104841880500317, 0.026928845793008804, 0.019565340131521225, 0.006846140138804913, 0.019197991117835045, -0.014971273951232433, -0.022601334378123283, -0.05624077096581459, -0.031603459268808365, 0.0005175537080504, -0.048893168568611145, -...
The Information-Theoretic Requirements of Subspace Clustering with Missing Data
https://proceedings.mlr.press/v48/pimentel-alarcon16.html
[ "Daniel Pimentel-Alarcon", "Robert Nowak" ]
null
null
Subspace clustering with missing data (SCMD) is a useful tool for analyzing incomplete datasets. Let d be the ambient dimension, and r the dimension of the subspaces. Existing theory shows that Nk = O(r d) columns per subspace are necessary for SCMD, and Nk =O(min d^(log d), d^(r+1) ) are sufficient. We close this gap,...
[]
null
85
null
null
[ -0.024674350395798683, -0.029353288933634758, 0.0025257214438170195, 0.0656554326415062, 0.07092441618442535, 0.015406838618218899, 0.018551746383309364, -0.004251563455909491, -0.023370830342173576, -0.02653205767273903, -0.03256555274128914, 0.008364219218492508, -0.052436769008636475, 0...
Online Learning with Feedback Graphs Without the Graphs
https://proceedings.mlr.press/v48/cohena16.html
[ "Alon Cohen", "Tamir Hazan", "Tomer Koren" ]
null
null
We study an online learning framework introduced by Mannor and Shamir (2011) in which the feedback is specified by a graph, in a setting where the graph may vary from round to round and is \emphnever fully revealed to the learner. We show a large gap between the adversarial and the stochastic cases. In the adversarial ...
[]
null
86
1605.07018
title_snapshot
[ -0.000604204717092216, -0.01642044261097908, -0.005022373981773853, 0.058440353721380234, 0.02330554835498333, 0.034043315798044205, 0.01830323226749897, 0.022342747077345848, -0.009424098767340183, -0.033746395260095596, -0.007489263080060482, -0.0030474141240119934, -0.0760127454996109, ...
PAC learning of Probabilistic Automaton based on the Method of Moments
https://proceedings.mlr.press/v48/glaude16.html
[ "Hadrien Glaude", "Olivier Pietquin" ]
null
null
Probabilitic Finite Automata (PFA) are generative graphical models that define distributions with latent variables over finite sequences of symbols, a.k.a. stochastic languages. Traditionally, unsupervised learning of PFA is performed through algorithms that iteratively improves the likelihood like the Expectation-Maxi...
[]
null
87
null
null
[ -0.01800478622317314, 0.011258412152528763, -0.026680422946810722, 0.024579515680670738, 0.033294614404439926, 0.042922161519527435, 0.01718827895820141, 0.006071207579225302, -0.017228106036782265, -0.03444648161530495, 0.013963945209980011, -0.017865272238850594, -0.07746686041355133, -0...
Estimating Structured Vector Autoregressive Models
https://proceedings.mlr.press/v48/melnyk16.html
[ "Igor Melnyk", "Arindam Banerjee" ]
null
null
While considerable advances have been made in estimating high-dimensional structured models from independent data using Lasso-type models, limited progress has been made for settings when the samples are dependent. We consider estimating structured VAR (vector auto-regressive model), where the structure can be captured...
[]
null
88
1602.06606
title_judge
[ 0.012109745293855667, -0.00820083450525999, 0.013856093399226665, -0.009553167968988419, 0.03362599387764931, 0.040240246802568436, 0.05874348804354668, -0.006572946906089783, -0.03730328753590584, -0.046612586826086044, 0.023699043318629265, -0.005701717920601368, -0.06707244366407394, -0...
Mixing Rates for the Alternating Gibbs Sampler over Restricted Boltzmann Machines and Friends
https://proceedings.mlr.press/v48/tosh16.html
[ "Christopher Tosh" ]
null
null
Alternating Gibbs sampling is a modification of classical Gibbs sampling where several variables are simultaneously sampled from their joint conditional distribution. In this work, we investigate the mixing rate of alternating Gibbs sampling with a particular emphasis on Restricted Boltzmann Machines (RBMs) and variant...
[]
null
89
null
null
[ -0.006340349093079567, -0.023102890700101852, -0.026723023504018784, 0.003462445456534624, 0.01652214303612709, 0.00613486347720027, 0.03564966470003128, -0.02134050242602825, -0.0381958894431591, -0.060485292226076126, -0.0017113623907789588, -0.025225814431905746, -0.05148172006011009, -...
Polynomial Networks and Factorization Machines: New Insights and Efficient Training Algorithms
https://proceedings.mlr.press/v48/blondel16.html
[ "Mathieu Blondel", "Masakazu Ishihata", "Akinori Fujino", "Naonori Ueda" ]
null
null
Polynomial networks and factorization machines are two recently-proposed models that can efficiently use feature interactions in classification and regression tasks. In this paper, we revisit both models from a unified perspective. Based on this new view, we study the properties of both models and propose new efficient...
[]
null
90
1607.08810
title_snapshot
[ -0.02012544311583042, -0.029425786808133125, -0.005989161320030689, 0.030233917757868767, 0.04107101634144783, 0.028933517634868622, 0.0144352400675416, -0.012422081083059311, -0.011319370940327644, -0.03889504447579384, -0.026104092597961426, 0.010944634675979614, -0.047468479722738266, 0...
A New PAC-Bayesian Perspective on Domain Adaptation
https://proceedings.mlr.press/v48/germain16.html
[ "Pascal Germain", "Amaury Habrard", "François Laviolette", "Emilie Morvant" ]
null
null
We study the issue of PAC-Bayesian domain adaptation: We want to learn, from a source domain, a majority vote model dedicated to a target one. Our theoretical contribution brings a new perspective by deriving an upper-bound on the target risk where the distributions’ divergence - expressed as a ratio - controls the tra...
[]
null
91
1506.04573
title_snapshot
[ -0.018637387081980705, -0.007819152437150478, 0.004695044364780188, 0.04667121171951294, 0.03491416573524475, 0.019199006259441376, 0.04309763014316559, -0.03141533583402634, -0.008705368265509605, -0.022976549342274666, -0.024696938693523407, 0.023516952991485596, -0.08564586192369461, 0....
Correlation Clustering and Biclustering with Locally Bounded Errors
https://proceedings.mlr.press/v48/puleo16.html
[ "Gregory Puleo", "Olgica Milenkovic" ]
null
null
We consider a generalized version of the correlation clustering problem, defined as follows. Given a complete graph G whose edges are labeled with + or -, we wish to partition the graph into clusters while trying to avoid errors: + edges between clusters or - edges within clusters. Classically, one seeks to minimize th...
[]
null
92
1506.08189
title_snapshot
[ 0.012943045236170292, 0.026239778846502304, 0.002516039414331317, 0.044444579631090164, 0.03860723227262497, 0.043068692088127136, 0.009871674701571465, 0.005188473500311375, -0.016124743968248367, -0.024259135127067566, -0.010845639742910862, -0.04232832416892052, -0.08381106704473495, 0....
PAC Lower Bounds and Efficient Algorithms for The Max K-Armed Bandit Problem
https://proceedings.mlr.press/v48/david16.html
[ "Yahel David", "Nahum Shimkin" ]
null
null
We consider the Max K-Armed Bandit problem, where a learning agent is faced with several stochastic arms, each a source of i.i.d. rewards of unknown distribution. At each time step the agent chooses an arm, and observes the reward of the obtained sample. Each sample is considered here as a separate item with the reward...
[]
null
93
1512.07650
title_judge
[ -0.03215845301747322, -0.006525794509798288, -0.014323730021715164, 0.0490434467792511, 0.04899973049759865, 0.013500162400305271, 0.03186074271798134, -0.012521116062998772, -0.019631097093224525, -0.03634483739733696, -0.014617708511650562, -0.006908652372658253, -0.04926853999495506, -0...
A Comparative Analysis and Study of Multiview CNN Models for Joint Object Categorization and Pose Estimation
https://proceedings.mlr.press/v48/elhoseiny16.html
[ "Mohamed Elhoseiny", "Tarek El-Gaaly", "Amr Bakry", "Ahmed Elgammal" ]
null
null
In the Object Recognition task, there exists a dichotomy between the categorization of objects and estimating object pose, where the former necessitates a view-invariant representation, while the latter requires a representation capable of capturing pose information over different categories of objects. With the rise o...
[]
null
94
1511.05175
title_judge
[ 0.015349348075687885, -0.009118826128542423, -0.04045860469341278, 0.041111744940280914, 0.01683763600885868, 0.03382273390889168, -0.001972984755411744, 0.004778845235705376, -0.027591757476329803, -0.046794064342975616, -0.027039891108870506, -0.013238769955933094, -0.08462442457675934, ...
BASC: Applying Bayesian Optimization to the Search for Global Minima on Potential Energy Surfaces
https://proceedings.mlr.press/v48/carr16.html
[ "Shane Carr", "Roman Garnett", "Cynthia Lo" ]
null
null
We present a novel application of Bayesian optimization to the field of surface science: rapidly and accurately searching for the global minimum on potential energy surfaces. Controlling molecule-surface interactions is key for applications ranging from environmental catalysis to gas sensing. We present pragmatic techn...
[]
null
95
null
null
[ 0.0030386466532945633, 0.04924307018518448, -0.0012981032487004995, -0.009155666455626488, 0.02669432759284973, -0.014276054687798023, 0.025966022163629532, -0.013702807947993279, 0.013031008653342724, -0.06354827433824539, 0.029922502115368843, 0.028992293402552605, -0.0344524160027504, -...
On the Iteration Complexity of Oblivious First-Order Optimization Algorithms
https://proceedings.mlr.press/v48/arjevani16.html
[ "Yossi Arjevani", "Ohad Shamir" ]
null
null
We consider a broad class of first-order optimization algorithms which are \emphoblivious, in the sense that their step sizes are scheduled regardless of the function under consideration, except for limited side-information such as smoothness or strong convexity parameters. With the knowledge of these two parameters, w...
[]
null
96
1605.03529
title_snapshot
[ -0.06308626383543015, -0.0034785261377692223, 0.01420209277421236, 0.03759218379855156, 0.03361128643155098, 0.04485708102583885, 0.032644402235746384, 0.004148161970078945, -0.013376177288591862, -0.01921989768743515, -0.013884581625461578, -0.024221694096922874, -0.054503146559000015, -0...
Stochastic Variance Reduced Optimization for Nonconvex Sparse Learning
https://proceedings.mlr.press/v48/lid16.html
[ "Xingguo Li", "Tuo Zhao", "Raman Arora", "Han Liu", "Jarvis Haupt" ]
null
null
We propose a stochastic variance reduced optimization algorithm for solving a class of large-scale nonconvex optimization problems with cardinality constraints, and provide sufficient conditions under which the proposed algorithm enjoys strong linear convergence guarantees and optimal estimation accuracy in high dimens...
[]
null
97
null
null
[ -0.011159607209265232, -0.008160650730133057, -0.0020636217668652534, 0.021360449492931366, 0.019457252696156502, 0.0399806909263134, 0.03450638800859451, 0.016085322946310043, -0.04489675164222717, -0.04001840576529503, 0.005273884627968073, -0.020451173186302185, -0.05349896848201752, 0....
Analysis of Variational Bayesian Factorizations for Sparse and Low-Rank Estimation
https://proceedings.mlr.press/v48/wipf16.html
[ "David Wipf" ]
null
null
Variational Bayesian (VB) approximations anchor a wide variety of probabilistic models, where tractable posterior inference is almost never possible. Typically based on the so-called VB mean-field approximation to the Kullback-Leibler divergence, a posterior distribution is sought that factorizes across groups of laten...
[]
null
98
null
null
[ -0.01024462841451168, 0.0064158872701227665, 0.014017270877957344, 0.025265196338295937, 0.026450371369719505, 0.03448517620563507, 0.03300239518284798, -0.022883644327521324, -0.05254816263914108, -0.055063795298337936, -0.019590774551033974, 0.007691520266234875, -0.06221722811460495, 0....
Fast k-means with accurate bounds
https://proceedings.mlr.press/v48/newling16.html
[ "James Newling", "Francois Fleuret" ]
null
null
We propose a novel accelerated exact k-means algorithm, which outperforms the current state-of-the-art low-dimensional algorithm in 18 of 22 experiments, running up to 3 times faster. We also propose a general improvement of existing state-of-the-art accelerated exact k-means algorithms through better estimates of the ...
[]
null
99
1602.02514
title_snapshot
[ -0.02290412411093712, -0.021561406552791595, 0.02699575014412403, 0.016425305977463722, 0.033621691167354584, 0.04183114692568779, 0.03773164749145508, 0.008467597886919975, 0.0014423237880691886, -0.027338551357388496, -0.011835714802145958, -0.028989410027861595, -0.0524996779859066, 0.0...
Boolean Matrix Factorization and Noisy Completion via Message Passing
https://proceedings.mlr.press/v48/ravanbakhsha16.html
[ "Siamak Ravanbakhsh", "Barnabas Poczos", "Russell Greiner" ]
null
null
Boolean matrix factorization and Boolean matrix completion from noisy observations are desirable unsupervised data-analysis methods due to their interpretability, but hard to perform due to their NP-hardness. We treat these problems as maximum a posteriori inference problems in a graphical model and present a message p...
[]
null
100
1509.08535
title_snapshot
[ -0.007976189255714417, -0.014334979467093945, 0.004101468715816736, 0.017017027363181114, 0.0529552660882473, 0.0020487648434937, 0.027024295181035995, 0.0009951998945325613, -0.03407057747244835, -0.05267917364835739, -0.015385923907160759, 0.004899709485471249, -0.05991994962096214, 0.00...