Unnamed: 0
int64
0
7.24k
id
int64
1
7.28k
raw_text
stringlengths
9
124k
vw_text
stringlengths
12
15k
5,800
6,248
Wasserstein Training of Restricted Boltzmann Machines Gr?goire Montavon Technische Universit?t Berlin Klaus-Robert M?ller? Technische Universit?t Berlin gregoire.montavon@tu-berlin.de klaus-robert.mueller@tu-berlin.de Marco Cuturi CREST, ENSAE, Universit? Paris-Saclay marco.cuturi@ensae.fr Abstract Boltzmann machi...
6248 |@word version:3 stronger:2 villani:1 distribue:1 twelfth:1 heuristically:2 decomposition:2 covariance:1 contrastive:1 mention:1 reduction:1 initial:1 configuration:1 united:1 document:1 interestingly:1 rightmost:2 freitas:1 recovered:1 must:2 readily:1 written:1 visible:2 partition:2 shape:2 plot:5 update:1 gener...
5,801
6,249
Human Decision-Making under Limited Time Pedro A. Ortega Department of Psychology University of Pennsylvania Philadelphia, PA 19104 ope@seas.upenn.edu Alan A. Stocker Department of Psychology University of Pennsylvania Philadelphia, PA 19014 astocker@sas.upenn.edu Abstract Subjective expected utility theory assumes ...
6249 |@word trial:24 determinant:1 cingulate:2 version:8 inversion:1 logit:1 integrative:1 gradual:3 crucially:2 decomposition:14 attainable:2 pick:2 pressure:3 paid:1 thereby:1 united:1 n000141110744:1 subjective:8 savage:1 anterior:2 conjunctive:1 john:1 plot:4 update:1 depict:1 implying:2 cue:2 device:1 accordingly:...
5,802
625
Visual Motion Computation in Analog VLSI using Pulses Rahul Sarpeshkar, Wyeth Bair and Christof Koch Computation and Neural Systems Program California Institute of Technology Pasadena, CA 91125. Abstract The real time computation of motion from real images using a single chip with integrated sensors is a hard problem...
625 |@word version:3 rising:2 pulse:46 excited:2 harder:1 series:1 t7:1 tuned:3 zerocrossings:2 current:2 com:1 must:1 motor:4 designed:1 progressively:1 v:6 half:4 lr:1 filtered:2 detecting:6 node:1 location:8 height:1 prove:1 resistive:1 behavior:2 inspired:2 correlator:1 window:2 increasing:1 begin:1 provided:1 circ...
5,803
6,250
Pruning Random Forests for Prediction on a Budget Feng Nan Systems Engineering Boston University fnan@bu.edu Joseph Wang Electrical Engineering Boston University joewang@bu.edu Venkatesh Saligrama Electrical Engineering Boston University srv@bu.edu Abstract We propose to prune a random forest (RF) for resource-cons...
6250 |@word repository:1 middle:1 polynomial:2 nd:1 seek:3 incurs:2 harder:1 reduction:2 initial:1 liu:1 contains:3 sherali:1 offering:1 ours:1 document:6 greedymiser:4 outperforms:5 existing:4 err:2 current:1 z2:2 nt:1 com:1 yet:2 must:2 realize:2 subsequent:2 happen:1 remove:1 plot:2 update:2 depict:1 v:2 greedy:4 le...
5,804
6,251
Learning Sensor Multiplexing Design through Back-propagation Ayan Chakrabarti Toyota Technological Institute at Chicago 6045 S. Kenwood Ave., Chicago, IL ayanc@ttic.edu Abstract Recent progress on many imaging and vision tasks has been driven by the use of deep feed-forward neural networks, which are trained by propa...
6251 |@word version:5 inversion:1 compression:2 seek:2 tried:1 rgb:8 sgd:2 shot:1 carry:2 initial:2 liu:1 contains:1 interestingly:1 outperforms:2 existing:2 past:1 current:1 z2:1 comparing:1 yet:1 gpu:2 chicago:2 visible:2 informative:1 blur:2 periodically:1 cheap:1 designed:1 treating:1 drop:1 update:2 half:1 fewer:1...
5,805
6,252
Optimal Sparse Linear Encoders and Sparse PCA Malik Magdon-Ismail Rensselaer Polytechnic Institute, Troy, NY 12211 magdon@cs.rpi.edu Christos Boutsidis New York, NY christos.boutsidis@gmail.com Abstract Principal components analysis (PCA) is the optimal linear encoder of data. Sparse linear encoders (e.g., sparse PC...
6252 |@word madelon:1 middle:1 version:3 polynomial:3 norm:4 stronger:1 loading:3 r13:1 underline:1 kbkf:1 open:1 d2:1 calculus:1 seems:1 nscta:1 sammon:1 decomposition:2 p0:1 q1:10 pick:1 tr:5 sepulchre:1 reduction:5 contains:1 dspca:1 ours:1 document:1 existing:4 err:3 ka:1 com:1 current:1 rpi:1 gmail:1 yet:1 must:1 ...
5,806
6,253
Blazing the trails before beating the path: Sample-efficient Monte-Carlo planning Jean-Bastien Grill Michal Valko SequeL team, INRIA Lille - Nord Europe, France jean-bastien.grill@inria.fr michal.valko@inria.fr R?mi Munos Google DeepMind, UK? munos@google.com Abstract You are a robot and you live in a Markov decisi...
6253 |@word version:2 eliminating:1 polynomial:9 open:4 r:3 pick:2 harder:1 initial:1 contains:1 selecting:2 daniel:1 ours:1 current:1 com:1 michal:2 assigning:1 guez:2 john:2 belmont:1 numerical:1 enables:2 cheap:1 update:1 generative:10 intelligence:4 vanishing:2 aja:1 provides:2 characterization:1 node:94 philipp:1 ...
5,807
6,254
Domain Separation Networks Konstantinos Bousmalis? Google Brain Mountain View, CA konstantinos@google.com Nathan Silberman Google Research New York, NY nsilberman@google.com George Trigeorgis? ? Imperial College London London, UK g.trigeorgis@imperial.ac.uk Dilip Krishnan Google Research Cambridge, MA dilipkay@googl...
6254 |@word private:27 version:3 cnn:5 norm:3 open:2 cleanly:1 covariance:1 brightness:1 lepetit:1 moment:1 initial:1 contains:3 efficacy:1 salzmann:3 ours:4 document:1 outperforms:4 existing:4 bitmap:1 com:5 nt:3 comparing:1 cad:1 si:5 realistic:2 additive:1 blur:1 shape:3 hypothesize:1 remove:1 v:1 alone:1 leaf:1 bis...
5,808
6,255
Conditional Generative Moment-Matching Networks Yong Ren, Jialian Li, Yucen Luo, Jun Zhu? Dept. of Comp. Sci. & Tech., TNList Lab; Center for Bio-Inspired Computing Research State Key Lab for Intell. Tech. & Systems, Tsinghua University, Beijing, China {renyong15, luoyc15, jl12}@mails.tsinghua.edu.cn; dcszj@tsinghua.ed...
6255 |@word cnn:11 version:3 briefly:1 middle:4 norm:4 hu:1 tr:9 harder:1 tnlist:1 ld:4 moment:7 configuration:1 series:2 contains:1 practiced:1 rkhs:9 outperforms:1 contextual:4 od:1 luo:1 comparing:1 exy:1 universality:1 activation:2 written:1 concatenate:1 subsequent:1 hofmann:1 laplacianfaces:1 update:1 v:1 generat...
5,809
6,256
A Credit Assignment Compiler for Joint Prediction Kai-Wei Chang University of Virginia kw@kwchang.net He He University of Maryland hhe@cs.umd.edu John Langford Microsoft Research jcl@microsoft.com Hal Daum? III University of Maryland me@hal3.name Stephane Ross Google stephaneross@google.com Abstract Many machine ...
6256 |@word middle:3 version:1 tadepalli:2 termination:4 tried:1 accounting:1 sgd:8 yih:1 reduction:4 contains:1 score:1 tuned:4 ours:1 rightmost:1 existing:4 bradley:1 current:7 com:3 comparing:1 beygelzimer:1 si:3 must:2 parsing:8 john:1 exposing:1 hofmann:1 drop:1 designed:1 update:7 v:1 implying:1 half:1 greedy:1 a...
5,810
6,257
Joint Line Segmentation and Transcription for End-to-End Handwritten Paragraph Recognition Th?odore Bluche A2iA SAS 39 rue de la Bienfaisance 75008 Paris tb@a2ia.com Abstract Offline handwriting recognition systems require cropped text line images for both training and recognition. On the one hand, the annotation of ...
6257 |@word torsten:1 version:1 laurence:1 hector:2 open:1 tried:1 mdlstm:20 ndez:1 contains:2 score:2 document:27 interestingly:1 outperforms:2 existing:1 current:2 com:1 comparing:1 michal:2 activation:2 yet:1 subsequent:1 happen:1 romero:2 christian:2 designed:1 interpretable:1 progressively:1 intelligence:1 prohibi...
5,811
6,258
Incremental Boosting Convolutional Neural Network for Facial Action Unit Recognition Shizhong Han, Zibo Meng, Ahmed Shehab Khan, Yan Tong Department of Computer Science & Engineering, University of South Carolina, Columbia, SC {han38, mengz, akhan}@email.sc.edu, tongy@cse.sc.edu Abstract Recognizing facial action uni...
6258 |@word multitask:1 cnn:154 open:1 simulation:1 carolina:1 contraction:1 hager:1 liu:6 contains:2 exclusively:1 selecting:2 score:25 tuned:2 outperforms:4 current:7 guadarrama:1 activation:13 attracted:1 shape:2 designed:2 drop:1 update:2 v:2 half:1 selected:19 plane:2 incredible:1 ith:1 short:1 detecting:1 boostin...
5,812
6,259
Variational Bayes on Monte Carlo Steroids Aditya Grover, Stefano Ermon Department of Computer Science Stanford University {adityag,ermon}@cs.stanford.edu Abstract Variational approaches are often used to approximate intractable posteriors or normalization constants in hierarchical latent variable models. While often ...
6259 |@word mild:1 version:1 contrastive:1 q1:1 sgd:2 reduction:1 moment:2 configuration:6 score:1 uma:1 document:10 subjective:1 existing:1 activation:1 additive:1 partition:2 visible:3 update:5 hash:4 generative:14 half:1 advancement:2 devising:1 blei:3 node:1 wierstra:1 qualitative:3 consists:4 combine:1 introduce:2...
5,813
626
A Connectionist Symbol Manipulator That Discovers the Structure of Context-Free Languages Michael C. Mozer and Sreerupa Das Department of Computer Science & Institute of Cognitive Science University of Colorado Boulder, CO 80309-0430 Abstract We present a neural net architecture that can discover hierarchical and rec...
626 |@word briefly:1 version:2 r:1 propagate:1 shading:1 reduction:2 initial:2 contains:2 current:2 activation:1 must:3 readily:1 parsing:2 written:2 numerical:2 designed:2 interpretable:1 half:1 selected:1 item:11 lr:6 compo:1 preference:1 attack:1 height:1 consists:1 theoretically:2 roughly:1 behavior:4 examine:1 con...
5,814
6,260
x2 x1 100 50 40 OKM OKM* OKM*+LPE OKM*+NPE 80 30 60 20 40 10 20 0 0 0.2 0.4 0.6 F?Score 0.8 1 0 0 OKM* OKM*+LPE OKM*+NPE 0.2 0.4 0.6 F?Score 0.8 1
6260 |@word lpe:2 score:2 okm:7 x2:1 npe:2 x1:1
5,815
6,261
Visual Question Answering with Question Representation Update (QRU) Ruiyu Li Jiaya Jia The Chinese University of Hong Kong {ryli,leojia}@cse.cuhk.edu.hk Abstract Our method aims at reasoning over natural language questions and visual images. Given a natural language question about an image, our model updates the ques...
6261 |@word kong:2 cnn:10 open:4 hu:2 shuicheng:1 seek:1 recursively:1 configuration:1 contains:4 score:1 selecting:1 hoiem:1 outperforms:3 activation:2 yet:1 gpu:1 visible:1 blur:1 informative:1 confirming:1 designed:2 drop:2 update:14 sukhbaatar:2 selected:1 ith:3 num:2 provides:1 cse:1 location:6 sits:2 five:1 heigh...
5,816
6,262
Adaptive Newton Method for Empirical Risk Minimization to Statistical Accuracy Aryan Mokhtari? University of Pennsylvania aryanm@seas.upenn.edu Hadi Daneshmand? ETH Zurich, Switzerland hadi.daneshmand@inf.ethz.ch Thomas Hofmann ETH Zurich, Switzerland thomas.hofmann@inf.ethz.ch Aurelien Lucchi ETH Zurich, Switzerlan...
6262 |@word mild:1 inversion:5 seems:1 stronger:1 consequential:1 norm:1 reused:1 nd:2 urb:1 citeseer:1 sgd:9 mention:1 reduction:3 initial:8 contains:3 united:2 outperforms:1 current:2 comparing:1 must:1 numerical:4 subsequent:2 kdd:1 hofmann:4 plot:2 update:6 juditsky:1 intelligence:2 website:1 item:1 amir:1 beginnin...
5,817
6,263
Learning Deep Parsimonious Representations Renjie Liao1 , Alexander Schwing2 , Richard S. Zemel1,3 , Raquel Urtasun1 University of Toronto1 University of Illinois at Urbana-Champaign2 Canadian Institute for Advanced Research3 {rjliao, zemel, urtasun}@cs.toronto.edu, aschwing@illinois.edu Abstract In this paper we aim...
6263 |@word trial:2 compression:3 norm:1 nd:4 propagate:1 decomposition:1 sgd:1 tr:1 shot:7 initial:1 liu:1 contains:2 score:1 denoting:1 document:1 outperforms:1 current:5 com:1 comparing:1 guadarrama:1 activation:3 assigning:1 partition:1 shape:2 enables:1 remove:1 drop:1 seeding:1 update:10 hash:1 cue:1 selected:1 w...
5,818
6,264
Dialog-based Language Learning Jason Weston Facebook AI Research, New York. jase@fb.com Abstract A long-term goal of machine learning research is to build an intelligent dialog agent. Most research in natural language understanding has focused on learning from fixed training sets of labeled data, with supervision eit...
6264 |@word open:2 instruction:1 r:1 crucially:1 tried:1 harder:2 carry:1 initial:1 score:1 selecting:1 past:1 existing:2 o2:3 com:1 comparing:1 surprising:2 must:2 parsing:2 john:16 realistic:2 subsequent:3 informative:1 ronan:1 pertinent:1 cracking:1 succeeding:1 v:1 infant:1 intelligence:1 selected:2 guess:1 sukhbaa...
5,819
6,265
A Sparse Interactive Model for Matrix Completion with Side Information Jin Lu Guannan Liang Jiangwen Sun Jinbo Bi University of Connecticut Storrs, CT 06269 {jin.lu, guannan.liang, jiangwen.sun, jinbo.bi}@uconn.edu Abstract Matrix completion methods can benefit from side information besides the partially observed mat...
6265 |@word kgk:2 version:2 norm:16 stronger:2 nd:1 d2:4 hu:1 simulation:4 linearized:4 heiser:1 decomposition:2 hsieh:2 liu:2 kpv:1 tuned:1 outperforms:1 existing:5 kmk:1 recovered:5 jinbo:3 reaction:1 written:1 reminiscent:1 informative:4 update:1 rd2:1 selected:1 ith:1 chiang:2 math:1 location:1 simpler:1 zhang:1 ma...
5,820
6,266
Ancestral Causal Inference Sara Magliacane VU Amsterdam & University of Amsterdam sara.magliacane@gmail.com Tom Claassen Radboud University Nijmegen tomc@cs.ru.nl Joris M. Mooij University of Amsterdam j.m.mooij@uva.nl Abstract Constraint-based causal discovery from limited data is a notoriously difficult challenge ...
6266 |@word mild:3 version:8 middle:1 achievable:1 seems:2 open:1 calculus:1 closure:1 willing:1 crucially:2 cyclic:2 series:2 score:15 contains:1 bootstrapped:16 interestingly:2 outperforms:1 existing:2 ramsey:1 recovered:1 com:2 comparing:1 plcg:19 gmail:1 assigning:1 happen:1 enables:1 plot:4 v:1 greedy:1 discoverin...
5,821
6,267
Regularized Nonlinear Acceleration Damien Scieur INRIA & D.I., UMR 8548, ?cole Normale Sup?rieure, Paris, France. damien.scieur@inria.fr Alexandre d?Aspremont CNRS & D.I., UMR 8548, ?cole Normale Sup?rieure, Paris, France. aspremon@di.ens.fr Francis Bach INRIA & D.I., UMR 8548, ?cole Normale Sup?rieure, Paris, Franc...
6267 |@word madelon:3 version:3 polynomial:31 norm:7 km:1 recursively:1 moment:1 reduction:1 initial:1 series:3 daniel:1 kx0:15 current:1 ka:6 comparing:1 must:1 written:3 stemming:1 numerical:9 zaid:1 extrapolating:1 interpretable:1 update:1 implying:1 stationary:1 device:1 xk:7 smith:2 recherche:1 ck2:4 num:1 iterate...
5,822
6,268
MetaGrad: Multiple Learning Rates in Online Learning Tim van Erven Leiden University tim@timvanerven.nl Wouter M. Koolen Centrum Wiskunde & Informatica wmkoolen@cwi.nl Abstract In online convex optimization it is well known that certain subclasses of objective functions are much easier than arbitrary convex functions...
6268 |@word briefly:1 version:26 stronger:1 seems:3 norm:5 open:2 mehta:1 d2:11 simulation:2 seek:1 linearized:2 covariance:5 gradual:1 incurs:3 boundedness:1 series:3 contains:2 tuned:1 erven:9 existing:2 past:1 discretization:1 surprising:1 luo:2 intriguing:1 bd:1 tilted:3 partition:2 happen:1 shape:2 drop:1 update:8...
5,823
6,269
Graphical Time Warping for Joint Alignment of Multiple Curves Yizhi Wang Virginia Tech yzwang@vt.edu David J. Miller Pennsylvania State University djmiller@engr.psu.edu Yue Wang Virginia Tech yuewang@vt.edu Kira Poskanzer University of California, San Francisco Kira.Poskanzer@ucsf.edu Lin Tian University of Califo...
6269 |@word middle:2 advantageous:1 seems:1 simulation:6 paid:1 fortuitous:1 necessity:1 series:23 contains:1 liquid:1 denoting:1 tuned:1 interestingly:2 batista:1 outperforms:1 existing:3 comparing:1 yet:1 must:1 readily:3 written:1 distant:1 additive:1 subsequent:1 shape:2 designed:1 progressively:1 v:2 intelligence:...
5,824
627
History-dependent Attractor Neural Networks Isaac Meilijson Eytan Ruppin School of Mathematical Sciences Raymond and Beverly Sackler Faculty of Exact Sciences Tel-A viv University, 69978 Tel-Aviv, Israel. Abstract We present a methodological framework enabling a detailed description of the performance of Hopfield-like...
627 |@word trial:1 faculty:1 open:1 simulation:2 moment:1 initial:3 selecting:1 past:1 current:4 activation:14 si:1 dx:1 afl:1 numerical:2 plot:1 update:1 signalling:5 contribute:1 sigmoidal:4 lor:1 mathematical:2 become:2 retrieving:1 consists:1 xji:1 brain:2 actual:2 becomes:1 israel:1 what:1 quantitative:1 every:1 u...
5,825
6,270
Stochastic Multiple Choice Learning for Training Diverse Deep Ensembles Stefan Lee Virginia Tech steflee@vt.edu Senthil Purushwalkam Carnegie Mellon University spurushw@andrew.cmu.edu David Crandall Indiana University djcran@indiana.edu Michael Cogswell Virginia Tech cogswell@vt.edu Viresh Ranjan Virginia Tech rvir...
6270 |@word kohli:2 cnn:14 faculty:3 version:1 middle:1 retraining:5 everingham:1 open:1 surfboard:6 seek:1 propagate:1 prasad:1 pick:2 sgd:4 rivera:5 accommodate:1 reduction:1 initial:1 liu:1 efficacy:3 score:3 selecting:1 tuned:3 past:1 existing:8 outperforms:5 com:2 assigning:1 gpu:2 neuraltalk2:1 distant:1 partitio...
5,826
6,271
Global Optimality of Local Search for Low Rank Matrix Recovery Srinadh Bhojanapalli srinadh@ttic.edu Behnam Neyshabur bneyshabur@ttic.edu Nathan Srebro nati@ttic.edu Toyota Technological Institute at Chicago Abstract We show that there are no spurious local minima in the non-convex factorized parametrization of low-...
6271 |@word mild:2 polynomial:4 stronger:1 norm:4 seems:3 nd:3 decomposition:3 sgd:5 sepulchre:1 necessity:1 liu:3 ours:2 past:1 ka:3 recovered:1 comparing:1 subsequent:1 chicago:1 cheap:1 hoping:1 plot:2 update:2 implying:1 stationary:6 parametrization:2 matrix1:1 zhang:1 mathematical:3 along:4 symposium:1 yuan:1 comb...
5,827
6,272
Preference Completion from Partial Rankings Suriya Gunasekar University of Texas, Austin, TX, USA suriya@utexas.edu Oluwasanmi Koyejo University of Illinois, Urbana-Champaign, IL, USA sanmi@illinois.edu Joydeep Ghosh University of Texas,Austin, TX, USA ghosh@ece.utexas.edu Abstract We propose a novel and efficient ...
6272 |@word version:1 pw:1 norm:10 mcrank:1 c0:3 d2:22 km:3 seek:3 grey:1 steck:1 decomposition:1 p0:1 harder:1 initial:1 substitution:2 liu:3 score:33 contains:1 denoting:1 tuned:1 bootstrapped:2 neeman:1 outperforms:2 existing:1 current:2 optim:2 incidence:2 ganti:1 activation:3 numerical:7 partition:1 kdd:1 remove:1...
5,828
6,273
Fundamental Limits of Budget-Fidelity Trade-off in Label Crowdsourcing Farshad Lahouti Electrical Engineering Department, California Institute of Technology lahouti@caltech.edu Babak Hassibi Electrical Engineering Department, California Institute of Technology hassibi@caltech.edu Abstract Digital crowdsourcing (CS) i...
6273 |@word achievable:2 compression:3 instrumental:1 dekel:1 tedious:1 seek:1 bn:1 paid:1 accommodate:1 series:1 karger:1 subjective:2 csn:1 current:2 comparing:2 incidence:9 numerical:2 enables:1 remove:1 designed:1 item:22 accordingly:1 ruvolo:1 ith:1 provides:5 completeness:1 contribute:1 node:1 org:1 zhang:1 along...
5,829
6,274
Generalized Correspondence-LDA Models (GC-LDA) for Identifying Functional Regions in the Brain Timothy N. Rubin SurveyMonkey Oluwasanmi Koyejo Univ. of Illinois, Urbana-Champaign Michael N. Jones Indiana University Tal Yarkoni University of Texas at Austin Abstract This paper presents Generalized Correspondence-LD...
6274 |@word proceeded:1 fusiform:1 middle:2 version:12 loading:1 adrian:1 decomposition:1 fabrice:1 pick:1 reduction:1 liu:1 contains:1 genetic:1 document:18 longitudinal:1 outperforms:1 existing:2 current:4 nt:1 activation:47 assigning:3 distant:2 analytic:3 motor:2 remove:1 atlas:4 interpretable:1 medial:2 toro:1 alo...
5,830
6,275
Ladder Variational Autoencoders Casper Kaae S?nderby? casperkaae@gmail.com Tapani Raiko? tapani.raiko@aalto.fi S?ren Kaae S?nderby? skaaesonderby@gmail.com Lars Maal?e? larsma@dtu.dk Ole Winther?,? olwi@dtu.dk Abstract Variational autoencoders are powerful models for unsupervised learning. However deep models with...
6275 |@word nd:1 d2:1 bn:5 shot:1 harder:1 recursively:5 contains:1 series:1 score:4 tuned:2 current:2 com:6 z2:4 nt:2 comparing:1 surprising:1 gmail:2 subsequent:1 wanted:1 plot:2 progressively:1 v:1 generative:44 parameterization:5 beginning:1 blei:1 provides:6 complication:1 five:6 wierstra:2 dn:2 bowman:1 mathemati...
5,831
6,276
Select-and-Sample for Spike-and-Slab Sparse Coding Abdul-Saboor Sheikh Technical University of Berlin, Germany, and Cluster of Excellence Hearing4all University of Oldenburg, Germany, and SAP Innovation Center Network, Berlin sheikh.abdulsaboor@gmail.com J?rg L?cke Research Center Neurosensory Science and Cluster of E...
6276 |@word neurophysiology:1 nd:3 d2:2 seek:1 decomposition:1 arti:2 eld:29 garrigues:1 initial:1 liu:1 contains:1 exclusively:1 oldenburg:2 selecting:2 outperforms:1 recovered:2 com:1 si:18 gmail:1 assigning:1 numerical:1 cant:1 shape:2 plasticity:1 drop:1 interpretable:2 update:1 generative:14 selected:4 half:1 cult...
5,832
6,277
Refined Lower Bounds for Adversarial Bandits S?bastien Gerchinovitz Institut de Math?matiques de Toulouse Universit? Toulouse 3 Paul Sabatier Toulouse, 31062, France sebastien.gerchinovitz@math.univ-toulouse.fr Tor Lattimore Department of Computing Science University of Alberta Edmonton, Canada tor.lattimore@gmail.com...
6277 |@word achievable:1 stronger:1 rigged:1 open:3 q1:8 incurs:1 bs01:1 tuned:2 erven:2 existing:5 com:1 z2:1 surprising:1 comparing:1 allenberg:2 gmail:1 must:5 subsequent:1 informative:1 benign:1 gerchinovitz:2 designed:1 selected:1 beginning:2 ith:1 recherche:1 provides:2 math:4 unbounded:1 symposium:1 prove:11 com...
5,833
6,278
CRF-CNN: Modeling Structured Information in Human Pose Estimation Xiao Chu The Chinese University of Hong Kong xchu@ee.cuhk.edu.hk Wanli Ouyang The Chinese University of Hong Kong wlouyang@ee.cuhk.edu.hk Hongsheng Li The Chinese University of Hong Kong hsli@ee.cuhk.edu.hk Xiaogang Wang The Chinese University of Hong...
6278 |@word kong:5 cnn:41 dalal:1 seems:1 paredes:1 everingham:1 triggs:1 heuristically:1 concise:3 configuration:6 contains:3 score:6 liu:1 loeliger:1 ours:3 romera:1 existing:2 activation:1 chu:6 pcp:4 parsing:1 partition:1 drop:1 update:1 greedy:1 leaf:2 half:1 generative:1 ith:4 vanishing:1 colored:1 provides:2 nod...
5,834
6,279
Natural-Parameter Networks: A Class of Probabilistic Neural Networks Hao Wang, Xingjian Shi, Dit-Yan Yeung Hong Kong University of Science and Technology {hwangaz,xshiab,dyyeung}@cse.ust.hk Abstract Neural networks (NN) have achieved state-of-the-art performance in various applications. Unfortunately in applications w...
6279 |@word kong:1 middle:2 version:5 seems:1 d2:1 sgd:2 moment:5 ndez:1 lightweight:1 bc:8 document:7 interestingly:1 outperforms:1 existing:1 wd:12 z2:3 od:29 activation:13 tackling:1 dx:7 ust:1 readily:1 bd:8 gpu:1 citeulike:11 subsequent:2 kdd:2 shape:1 enables:2 designed:3 drop:1 update:1 depict:1 treating:1 plot:...
5,835
628
Unsupervised Discrimination of Clustered Data via Optimization of Binary Information Gain Nicol N. Schraudolph Computer Science & Engr. Dept. University of California, San Diego La Jolla, CA 92093-0114 Terrence J. Sejnowski Computational Neurobiology Laboratory The Salk Institute for Biological Studies San Diego, CA ...
628 |@word eliminating:1 proportion:1 seek:1 covariance:2 solid:1 accommodate:1 tsejnowski:1 barney:3 carry:1 initial:6 substitution:1 existing:1 nowlan:1 scatter:2 must:1 reminiscent:1 happen:1 informative:3 remove:1 plot:5 concert:1 update:1 discrimination:11 half:3 greedy:1 ith:1 provides:2 node:29 preference:1 math...
5,836
6,280
DeepMath - Deep Sequence Models for Premise Selection Alexander A. Alemi ? Google Inc. alemi@google.com Geoffrey Irving ? Google Inc. geoffreyi@google.com Fran?ois Chollet ? Google Inc. fchollet@google.com Christian Szegedy ? Google Inc. szegedy@google.com Niklas Een ? Google Inc. een@google.com Josef Urban ?? Czech...
6280 |@word repository:1 version:3 cnn:19 stronger:2 calculus:1 underperform:1 heiser:1 propagate:1 essay:1 attainable:1 pick:1 accommodate:1 recursively:1 reduction:1 initial:1 series:2 score:3 pub:1 itp:13 tuned:1 outperforms:2 existing:2 ramsey:1 guadarrama:1 com:7 steiner:1 gmail:1 written:2 parsing:2 gpu:1 devin:1...
5,837
6,281
Optimizing Affinity-Based Binary Hashing Using Auxiliary Coordinates Ramin Raziperchikolaei EECS, University of California, Merced rraziperchikolaei@ucmerced.edu ? Carreira-Perpin? ? an Miguel A. EECS, University of California, Merced mcarreira-perpinan@ucmerced.edu Abstract In supervised binary hashing, one wants t...
6281 |@word mild:1 kulis:2 version:3 briefly:1 disk:1 perpin:1 tried:1 crucially:1 hsieh:1 solid:1 harder:1 liblinear:2 reduction:6 liu:2 contains:2 initial:1 interestingly:1 outperforms:4 existing:2 current:3 comparing:1 must:2 written:2 concatenate:1 numerical:1 shape:1 designed:1 gist:2 progressively:1 plot:3 hash:9...
5,838
6,282
Stochastic Gradient Geodesic MCMC Methods ? Chang Liu? , Jun Zhu? , Yang Song?? Dept. of Comp. Sci. & Tech., TNList Lab; Center for Bio-Inspired Computing Research ? State Key Lab for Intell. Tech. & Systems, Tsinghua University, Beijing, China ? Dept. of Physics, Tsinghua University, Beijing, China {chang-li14@mails...
6282 |@word briefly:1 version:4 changyou:3 advantageous:1 norm:1 nd:13 proportion:1 open:1 scalably:1 km:2 simulation:4 covariance:2 tnlist:1 liu:1 series:1 score:1 salzmann:1 ours:2 document:4 existing:1 jupp:1 yet:1 dx:1 john:3 numerical:1 shape:1 analytic:1 sdes:1 enables:1 kv1:1 stationary:4 generative:2 intelligen...
5,839
6,283
Deconvolving Feedback Loops in Recommender Systems Ayan Sinha Purdue University sinhayan@mit.edu David F. Gleich Purdue University dgleich@purdue.edu Karthik Ramani Purdue University ramani@purdue.edu Abstract Collaborative filtering is a popular technique to infer users? preferences on new content based on the col...
6283 |@word exploitation:3 briefly:2 version:2 adomavicius:1 seems:1 norm:3 open:1 hu:2 r:52 simulation:1 decomposition:5 eng:1 mcauley:1 initial:1 series:4 score:38 exclusively:1 interestingly:1 current:1 contextual:2 com:1 surprising:1 si:3 scatter:4 assigning:1 rpi:2 chu:1 subsequent:2 thrust:1 kdd:3 enables:3 asymp...
5,840
6,284
Latent Attention For If-Then Program Synthesis Xinyun Chen? Shanghai Jiao Tong University Chang Liu Richard Shin UC Berkeley Dawn Song Mingcheng Chen? UIUC Abstract Automatic translation from natural language descriptions into programs is a longstanding challenging problem. In this work, we consider a simple yet ...
6284 |@word middle:1 norm:5 instruction:1 shot:10 liu:1 contains:3 denoting:1 document:1 fa8750:1 outperforms:3 existing:6 past:1 com:9 unction:4 yet:1 parsing:8 concatenate:1 numerical:1 subsequent:1 informative:1 designed:2 update:1 sukhbaatar:1 fewer:1 website:3 device:1 kushman:1 short:2 provides:3 along:3 transduc...
5,841
6,285
A Multi-step Inertial Forward?Backward Splitting Method for Non-convex Optimization Jingwei Liang and Jalal M. Fadili Normandie Univ, ENSICAEN, CNRS, GREYC {Jingwei.Liang,Jalal.Fadili}@greyc.ensicaen.fr Gabriel Peyr? CNRS, DMA, ENS Paris Gabriel.Peyre@ens.fr Abstract We propose a multi-step inertial Forward?Backward...
6285 |@word version:1 norm:10 nd:2 open:1 calculus:1 decomposition:1 solid:3 boundedness:2 initial:1 outperforms:1 existing:2 comparing:1 discretization:1 optim:1 yet:1 pcp:4 written:1 numerical:5 additive:2 plot:2 xk:36 short:1 ojasiewicz:9 iterates:5 characterization:1 zhang:1 mathematical:5 along:3 become:1 differen...
5,842
6,286
Barzilai-Borwein Step Size for Stochastic Gradient Descent Conghui Tan The Chinese University of Hong Kong chtan@se.cuhk.edu.hk Shiqian Ma The Chinese University of Hong Kong sqma@se.cuhk.edu.hk Yu-Hong Dai Chinese Academy of Sciences, Beijing, China dyh@lsec.cc.ac.cn Yuqiu Qian The University of Hong Kong qyq79@con...
6286 |@word kong:4 briefly:1 norm:2 pick:2 sgd:66 solid:5 recursively:1 hager:1 reduction:2 initial:7 cyclic:1 efficacy:1 tuned:13 past:1 existing:2 outperforms:1 current:1 must:1 numerical:15 update:4 fund:1 selected:1 xk:14 lr:4 iterates:3 zhang:4 mathematical:1 along:2 prove:5 manner:1 introduce:1 x0:5 frequently:1 ...
5,843
6,287
Pairwise Choice Markov Chains Stephen Ragain Management Science & Engineering Stanford University Stanford, CA 94305 sragain@stanford.edu Johan Ugander Management Science & Engineering Stanford University Stanford, CA 94305 jugander@stanford.edu Abstract As datasets capturing human choices grow in richness and scale...
6287 |@word kohli:1 exploitation:1 faculty:1 version:1 seems:1 norm:1 logit:10 proportion:1 instrumental:1 open:4 sheffet:1 commute:1 asks:1 blade:1 reduction:2 necessity:1 substitution:1 contains:3 score:2 cyclic:4 selecting:3 efficacy:1 symphony:2 outperforms:2 past:1 bradley:4 mishra:1 com:1 assigning:1 must:1 serge...
5,844
6,288
Split LBI: An Iterative Regularization Path with Structural Sparsity Chendi Huang1 , Xinwei Sun1 , Jiechao Xiong1 , Yuan Yao2,1 Peking University, 2 Hong Kong University of Science and Technology {cdhuang, sxwxiaoxiaohehe, xiongjiechao}@pku.edu.cn, yuany@ust.hk 1 Abstract An iterative regularization path with structu...
6288 |@word h:1 kong:1 repository:1 version:2 norm:1 d2:2 linearized:2 rgb:1 decomposition:1 solid:2 initial:1 liu:1 series:2 score:1 denoting:3 outperforms:1 africa:2 recovered:1 comparing:1 osh:4 si:1 yet:1 ust:1 shape:1 drop:3 championship:3 alone:1 intelligence:1 rudin:1 xk:3 short:1 provides:1 coarse:1 boosting:1 ...
5,845
6,289
An Ensemble Diversity Approach to Supervised Binary Hashing ? Carreira-Perpin? ? an Miguel A. EECS, University of California, Merced mcarreira-perpinan@ucmerced.edu Ramin Raziperchikolaei EECS, University of California, Merced rraziperchikolaei@ucmerced.edu Abstract Binary hashing is a well-known approach for fast ap...
6289 |@word kulis:1 polynomial:3 proportion:2 seems:3 disk:1 open:1 confirms:1 perpin:1 bn:6 decomposition:3 hsieh:1 liblinear:2 bai:1 liu:2 contains:3 series:2 selecting:2 reduction:5 initial:1 document:1 bootstrapped:3 ours:1 outperforms:2 wd:1 yet:1 written:3 john:1 partition:1 happen:1 shape:1 designed:1 gist:1 upd...
5,846
629
Hidden Markov Models in Molecular Biology: New Algorithms and Applications Yves Chauvin t Net-ID, Inc. 8, Cathy Place Menlo Park, CA 94305 Pierre Baldi ? Jet Propulsion Laboratory California Institute of Technology Pasadena, CA 91109 Tim H lmkapiller Division of Biology California Institute of Technology Marcella A....
629 |@word private:1 briefly:2 mammal:1 phosphorylation:1 initial:2 fragment:1 egfr:5 terminus:1 ala:1 current:1 virus:1 si:1 must:1 cruz:1 remove:1 treating:1 plot:1 update:2 selected:1 beginning:1 tertiary:1 characterization:1 phylogenetic:2 mathematical:1 dn:1 constructed:1 ucsc:1 consists:2 baldi:10 recognizable:1 ...
5,847
6,290
Measuring the reliability of MCMC inference with bidirectional Monte Carlo Roger B. Grosse Department of Computer Science University of Toronto Siddharth Ancha Department of Computer Science University of Toronto Daniel M. Roy Department of Statistics University of Toronto Abstract Markov chain Monte Carlo (MCMC)...
6290 |@word mild:1 version:2 simulation:2 moment:3 initial:6 configuration:1 lightweight:1 series:2 uncovered:1 daniel:1 disallows:1 tuned:1 subjective:3 wd:20 comparing:1 com:1 must:1 written:1 dechter:1 realistic:2 partition:6 informative:2 update:5 mackey:2 intelligence:1 leaf:2 generative:1 isotropic:1 realizing:1 ...
5,848
6,291
Unsupervised Learning from Noisy Networks with Applications to Hi-C Data Bo Wang?1 , Junjie Zhu2 , Oana Ursu3 , Armin Pourshafeie4 , Serafim Batzoglou1 and Anshul Kundaje3,1 2 1 Department of Computer Science, Stanford University Department of Electrical Engineering, Stanford University 3 Department of Genetics, Stanf...
6291 |@word kong:1 version:1 norm:2 dekker:3 sex:1 termination:1 hu:1 zelnik:1 seek:2 serafim:1 uncovers:1 tr:8 mcauley:1 initial:2 liu:2 contains:1 score:1 selecting:1 uncovered:1 genetic:1 existing:1 current:1 com:1 od:1 si:5 activation:1 numerical:1 partition:2 treating:1 drop:5 update:2 aside:1 implying:1 selected:...
5,849
6,292
Linear Contextual Bandits with Knapsacks Shipra Agrawal? Nikhil R. Devanur? Abstract We consider the linear contextual bandit problem with resource consumption, in addition to reward generation. In each round, the outcome of pulling an arm is a reward as well as a vector of resource consumptions. The expected values...
6292 |@word trial:2 exploitation:2 version:1 norm:3 open:1 km:11 pick:5 profit:1 harder:1 reduction:1 liu:1 pt0:8 k1d:1 ours:1 current:1 contextual:27 com:1 comparing:1 wilkens:1 tackling:1 chu:1 midway:1 enables:1 sponsored:1 update:5 half:1 beginning:1 core:1 provides:1 along:4 constructed:2 become:1 focs:1 prove:2 c...
5,850
6,293
Variance Reduction in Stochastic Gradient Langevin Dynamics Avinava Dubey? , Sashank J. Reddi? , Barnab?as P?oczos, Alexander J. Smola, Eric P. Xing Department of Machine Learning Carnegie-Mellon University Pittsburgh, PA 15213 {akdubey, sjakkamr, bapoczos, alex, epxing}@cs.cmu.edu Sinead A. Williamson IROM/Statistics ...
6293 |@word repository:1 version:2 changyou:2 norm:1 h2t:5 pick:2 thereby:1 ld:13 reduction:14 initial:2 series:1 selecting:1 rightmost:1 outperforms:1 current:4 discretization:1 nt:3 numerical:1 cant:2 cheap:1 designed:3 plot:4 update:8 stationary:1 selected:2 item:1 complementing:1 slowing:1 accordingly:1 experiment3...
5,851
6,294
Safe Policy Improvement by Minimizing Robust Baseline Regret Marek Petrik University of New Hampshire mpetrik@cs.unh.edu Mohammad Ghavamzadeh Adobe Research & INRIA Lille ghavamza@adobe.com Yinlam Chow Stanford University ychow@stanford.edu Abstract An important problem in sequential decision-making under uncertain...
6294 |@word briefly:1 polynomial:3 norm:1 p0:3 profit:1 solid:1 initial:1 contains:1 outperforms:1 existing:3 current:4 com:1 yet:1 readily:2 realistic:1 happen:1 stationary:1 intelligence:2 website:1 provides:2 mannor:2 node:3 simpler:2 along:1 constructed:3 prove:5 consists:1 combine:1 manner:1 introduce:1 x0:4 indee...
5,852
6,295
Can Active Memory Replace Attention? ?ukasz Kaiser Google Brain lukaszkaiser@google.com Samy Bengio Google Brain bengio@google.com Abstract Several mechanisms to focus attention of a neural network on selected parts of its input or memory have been used successfully in deep learning models in recent years. Attention...
6295 |@word compression:3 seems:1 norm:1 open:1 d2:1 tried:1 p0:2 pick:1 versatile:1 shot:1 harder:1 carry:1 substitution:1 contains:1 score:12 selecting:1 jimenez:3 liu:3 tuned:1 o2:2 existing:2 steiner:1 com:3 culprit:1 diederik:1 gpu:29 parsing:2 ronald:1 subsequent:1 devin:1 shape:7 remove:1 designed:2 plot:2 depic...
5,853
6,296
Kronecker Determinantal Point Processes Zelda Mariet Massachusetts Institute of Technology Cambridge, MA 02139 zelda@csail.mit.edu Suvrit Sra Massachusetts Institute of Technology Cambridge, MA 02139 suvrit@mit.edu Abstract Determinantal Point Processes (DPPs) are probabilistic models over all subsets a ground set of...
6296 |@word determinant:2 polynomial:1 seems:1 confirms:1 prominence:1 decomposition:3 covariance:1 profit:1 nystr:3 tr:4 initial:4 liu:1 contains:1 series:1 document:2 precluding:1 ka:3 si:4 assigning:1 bd:2 must:1 determinantal:20 realistic:1 partition:2 happen:1 subsequent:1 enables:3 drop:1 update:25 v:1 greedy:1 i...
5,854
6,297
Consistent Estimation of Functions of Data Missing Non-Monotonically and Not at Random Ilya Shpitser Department of Computer Science Johns Hopkins University ilyas@cs.jhu.edu Abstract Missing records are a perennial problem in analysis of complex data of all types, when the target of inference is some function of the ...
6297 |@word mild:4 version:3 briefly:1 open:1 simulation:4 mcar:2 configuration:2 series:2 daniel:1 ours:1 longitudinal:3 horvitz:2 existing:3 nonmonotone:1 yet:1 john:1 subsequent:1 realistic:1 benign:1 resampling:2 aside:1 intelligence:1 parameterization:4 mccallum:1 record:5 characterization:1 parameterizations:3 no...
5,855
6,298
Scaling Memory-Augmented Neural Networks with Sparse Reads and Writes Jack W Rae? jwrae Jonathan J Hunt? jjhunt Greg Wayne gregwayne Tim Harley tharley Alex Graves gravesa Ivo Danihelka danihelka Andrew Senior andrewsenior Timothy P Lillicrap countzero Google DeepMind @google.com Abstract Neural networks augme...
6298 |@word armand:1 version:2 compression:1 bptt:4 open:1 tried:1 jacob:1 tr:5 shot:3 series:1 att:1 contains:3 daniel:1 reynolds:2 existing:1 current:1 com:1 comparing:1 surprising:1 si:2 written:2 must:1 john:1 additive:1 subsequent:1 ronan:1 designed:1 update:3 progressively:1 hash:2 sukhbaatar:1 cue:1 prohibitive:...
5,856
6,299
Breaking the Bandwidth Barrier: Geometrical Adaptive Entropy Estimation Weihao Gao?, Sewoong Oh?, and Pramod Viswanath? University of Illinois at Urbana-Champaign Urbana, IL 61801 {wgao9,swoh,pramodv}@illinois.edu Abstract Estimators of information theoretic measures such as entropy and mutual information are a basic...
6299 |@word mild:1 determinant:1 briefly:1 polynomial:5 reshef:2 heuristically:2 hu:1 simulation:3 covariance:2 substitution:1 series:1 liu:1 renewed:1 past:1 existing:2 outperforms:5 ka:1 universality:1 dx:1 readily:1 grassberger:3 fn:2 numerical:7 additive:1 informative:1 remove:1 intelligence:2 kandasamy:1 discoveri...
5,857
63
1 CONNECTIVITY VERSUS ENTROPY Yaser S. Abu-Mostafa California Institute of Technology Pasadena, CA 91125 ABSTRACT How does the connectivity of a neural network (number of synapses per neuron) relate to the complexity of the problems it can handle (measured by the entropy)? Switching theory would suggest no relation a...
63 |@word implemented:1 concept:1 version:1 normalized:1 hence:10 restate:1 diagonal:6 alp:1 during:1 implementing:1 everything:1 accommodate:1 oa:1 ln2:1 hill:1 biological:1 complete:1 hold:3 length:2 lof:1 purposely:2 must:6 written:1 equivalently:1 mostafa:4 substituting:2 iv1:1 relate:1 a2:1 veal:2 designed:3 jl:2 ...
5,858
630
Attractor Neural Networks with Local Inhibition: from Statistical Physics to a Digital Programmable Integrated Circuit E. Pasero Dipartimento di Elettronica Politecnico di Torino 1-10129 Torino, Italy R. Zecchina Dipartimento di Fisica Teorica e INFN Universita. di Torino 1-10125 Torino, Italy Abstract Networks with...
630 |@word luk:1 version:4 seems:2 jijsj:1 simulation:2 r:1 pulse:1 reduction:1 phy:1 configuration:5 initial:2 hereafter:1 selecting:1 ala:1 past:1 existing:1 clash:1 od:1 si:1 activation:2 must:2 additive:1 numerical:3 partition:1 stationary:3 selected:2 device:8 short:2 provides:1 rc:1 become:2 inside:1 indeed:1 exp...
5,859
6,300
Examples are not Enough, Learn to Criticize! Criticism for Interpretability Been Kim? Allen Institute for AI beenkim@csail.mit.edu Rajiv Khanna UT Austin rajivak@utexas.edu Oluwasanmi Koyejo UIUC sanmi@illinois.edu Abstract Example-based explanations are widely used in the effort to improve the interpretability of h...
6300 |@word determinant:3 polynomial:1 seitz:1 asks:2 reduction:1 contains:2 score:3 selecting:6 offering:1 ours:2 rkhs:2 document:3 tuned:1 subjective:2 existing:2 genetic:1 comparing:1 surprising:1 assigning:1 must:2 written:3 additive:2 kdd:2 remove:1 designed:2 gist:1 interpretable:6 v:1 alone:2 greedy:9 selected:8...
5,860
6,301
Large-Scale Price Optimization via Network Flow Shinji Ito NEC Corporation s-ito@me.jp.nec.com Ryohei Fujimaki NEC Corporation rfujimaki@nec-labs.com Abstract This paper deals with price optimization, which is to find the best pricing strategy that maximizes revenue or profit, on the basis of demand forecasting mode...
6301 |@word mild:1 middle:1 polynomial:4 norm:1 km:2 simulation:4 git:4 decomposition:1 profit:17 contains:1 pub:1 prescriptive:5 ours:1 existing:4 com:3 written:1 additive:3 kdd:1 cheap:1 enables:1 update:2 half:1 selected:1 intelligence:5 accordingly:1 core:2 oblique:1 provides:1 org:1 treed:1 dn:1 constructed:1 ryoh...
5,861
6,302
Low-Rank Regression with Tensor Responses Guillaume Rabusseau and Hachem Kadri Aix Marseille Univ, CNRS, LIF, Marseille, France {firstname.lastname}@lif.univ-mrs.fr Abstract This paper proposes an efficient algorithm (HOLRR) to handle regression tasks where the outputs have a tensor structure. We formulate the regres...
6302 |@word h:1 multitask:3 trial:2 version:4 middle:1 polynomial:2 norm:3 paredes:1 d2:9 rgb:2 decomposition:11 invoking:1 solid:1 reduction:1 liu:2 series:1 contains:3 rkhs:1 ours:1 romera:1 outperforms:2 existing:1 comparing:1 written:3 readily:2 additive:1 numerical:2 kdd:1 analytic:1 designed:1 update:1 rd2:1 v:2 ...
5,862
6,303
Architectural Complexity Measures of Recurrent Neural Networks Saizheng Zhang1,?, Yuhuai Wu2,? , Tong Che4 , Zhouhan Lin1 , Roland Memisevic1,5 , Ruslan Salakhutdinov3,5 and Yoshua Bengio1,5 1 MILA, Universit? de Montr?al, 2 University of Toronto, 3 Carnegie Mellon University, 4 Institut des Hautes ?tudes Scientifiques...
6303 |@word mild:4 repository:1 version:1 nchen:1 compression:1 seems:1 nd:1 km:2 p0:2 paid:1 solid:1 recursively:1 carry:2 necessity:1 cyclic:11 series:1 contains:2 subword:1 outperforms:1 current:2 comparing:1 com:1 surprising:1 skipping:2 activation:2 diederik:1 amjad:1 must:1 plot:1 v:1 bart:1 intelligence:1 beginn...
5,863
6,304
Convolutional Neural Fabrics Shreyas Saxena Jakob Verbeek INRIA Grenoble ? Laboratoire Jean Kuntzmann Abstract Despite the success of CNNs, selecting the optimal architecture for a given task remains an open problem. Instead of aiming to select a single optimal architecture, we propose a ?fabric? that embeds an expone...
6304 |@word cnn:11 middle:1 version:3 advantageous:1 kokkinos:1 tedious:1 open:1 heuristically:1 hu:1 propagate:1 sgd:1 accommodate:1 necessity:1 liu:3 contains:5 ndez:1 selecting:1 hoiem:1 ours:7 document:1 deconvolutional:2 past:1 recovered:2 activation:19 reminiscent:1 finest:1 parsing:1 shape:2 progressively:2 v:1 ...
5,864
6,305
Linear Feature Encoding for Reinforcement Learning Zhao Song, Ronald Parr? , Xuejun Liao, Lawrence Carin Department of Electrical and Computer Engineering ? Department of Computer Science Duke University, Durham, NC 27708, USA Abstract Feature construction is of vital importance in reinforcement learning, as the qual...
6305 |@word h:1 version:3 norm:2 termination:4 km:4 seek:1 simulation:2 tried:1 decomposition:1 dealer:1 thereby:1 wrapper:1 outperforms:2 existing:3 current:2 ka:1 surprising:1 si:6 reminiscent:1 must:2 written:1 john:1 ronald:1 numerical:1 enables:1 plot:1 update:5 aside:1 v:1 greedy:4 selected:2 half:1 shut:1 accord...
5,865
6,306
Online ICA: Understanding Global Dynamics of Nonconvex Optimization via Diffusion Processes Chris Junchi Li Zhaoran Wang Han Liu Department of Operations Research and Financial Engineering, Princeton University {junchil, zhaoran, hanliu}@princeton.edu Abstract Solving statistical learning problems often involves nonco...
6306 |@word polynomial:1 norm:1 c0:2 calculus:1 d2:1 decomposition:8 sgd:33 v2o:2 moment:1 initial:7 liu:12 luo:1 written:1 john:1 distant:1 informative:1 analytic:4 drop:1 update:3 clumping:1 stationary:8 intelligence:1 accordingly:1 xk:1 short:1 blei:1 characterization:5 coarse:1 provides:3 iterates:6 traverse:5 loca...
5,866
6,307
The Parallel Knowledge Gradient Method for Batch Bayesian Optimization Jian Wu, Peter I. Frazier Cornell University Ithaca, NY, 14853 {jw926, pf98}@cornell.edu Abstract In many applications of black-box optimization, one can evaluate multiple points simultaneously, e.g. when evaluating the performances of several dif...
6307 |@word cnn:5 briefly:1 version:2 exploitation:1 open:2 multipoint:1 underperform:1 zilinskas:1 simulation:1 covariance:2 dramatic:1 initial:5 configuration:1 ndez:1 liu:2 zij:10 tuned:1 ours:1 reine:1 past:1 outperforms:2 com:2 z2:3 discretization:2 comparing:1 must:3 additive:1 numerical:2 burdick:1 designed:1 up...
5,867
6,308
Anchor-Free Correlated Topic Modeling: Identifiability and Algorithm Kejun Huang? Xiao Fu? Nicholas D. Sidiropoulos Department of Electrical and Computer Engineering University of Minnesota Minneapolis, MN 55455, USA huang663@umn.edu xfu@umn.edu nikos@ece.umn.edu Abstract In topic modeling, many algorithms that guara...
6308 |@word mild:1 trial:2 determinant:5 version:3 manageable:1 polynomial:2 norm:1 middle:1 justice:1 simulation:1 tried:1 decomposition:10 eng:1 covariance:1 pick:3 thereby:1 moment:1 liu:3 contains:1 lightweight:1 score:3 series:2 document:22 suppressing:1 hottopixx:1 existing:1 surprising:1 yet:1 attracted:1 must:1...
5,868
6,309
A Constant-Factor Bi-Criteria Approximation Guarantee for k-means++ Dennis Wei IBM Research Yorktown Heights, NY 10598, USA dwei@us.ibm.com Abstract This paper studies the k-means++ algorithm for clustering as well as the class of D` sampling algorithms to which k-means++ belongs. It is shown that for any constant fac...
6309 |@word cu:28 version:1 polynomial:4 stronger:1 d2:1 confirms:1 hu:5 recursively:1 initial:2 uncovered:12 series:1 selecting:5 katoh:1 existing:5 current:3 com:1 discretization:1 assigning:1 must:1 numerical:1 subsequent:1 asymptote:1 seeding:1 update:1 pursued:1 selected:4 half:1 leaf:1 fewer:1 provides:3 location...
5,869
631
Input Reconstruction Reliability Estimation Dean A. Pomerleau School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213 Abstract This paper describes a technique called Input Reconstruction Reliability Estimation (IRRE) for determining the response reliability of a restricted class of multi-layer per...
631 |@word hippocampus:1 jacob:4 eng:1 solid:3 exclusively:1 selecting:2 existing:1 current:3 nowlan:1 activation:8 cottrell:4 subsequent:1 update:1 intelligence:2 fewer:1 inspection:1 provides:2 location:2 sigmoidal:1 rc:1 along:3 incorrect:2 baldi:4 expected:1 frequently:2 planning:1 multi:2 automatically:1 actual:11...
5,870
6,310
Phased LSTM: Accelerating Recurrent Network Training for Long or Event-based Sequences Daniel Neil, Michael Pfeiffer, and Shih-Chii Liu Institute of Neuroinformatics University of Zurich and ETH Zurich Zurich, Switzerland 8057 {dneil, pfeiffer, shih}@ini.uzh.ch Abstract Recurrent Neural Networks (RNNs) have become th...
6310 |@word cnn:5 middle:2 version:3 manageable:1 wco:2 retraining:1 bf:2 c0:2 open:13 linearized:1 bn:16 thereby:4 shading:1 carry:1 reduction:1 initial:2 liu:3 cyclic:1 configuration:1 contains:1 daniel:1 bc:2 outperforms:1 past:2 current:1 comparing:1 com:1 surprising:1 activation:1 must:1 gpu:1 fn:2 plasticity:2 dr...
5,871
6,311
Iterative Refinement of the Approximate Posterior for Directed Belief Networks R Devon Hjelm University of New Mexico and the Mind Research Network dhjelm@mrn.org Kyunghyun Cho Courant Institute & Center for Data Science, New York University kyunghyun.cho@nyu.edu Junyoung Chung University of Montreal junyoung.chung@umo...
6311 |@word middle:1 faculty:1 version:5 consequential:1 nd:1 adrian:1 r:1 simulation:1 sgd:1 reduction:2 initial:6 configuration:6 series:1 jimenez:1 tuned:1 fa8750:1 existing:1 freitas:1 current:1 com:2 comparing:2 si:2 yet:1 diederik:2 must:1 written:1 gpu:1 refines:1 shape:1 update:5 resampling:2 nebojsa:1 generati...
5,872
6,312
Mapping Estimation for Discrete Optimal Transport Micha?el Perrot Univ Lyon, UJM-Saint-Etienne, CNRS, Lab. Hubert Curien UMR 5516, F-42023 michael.perrot@univ-st-etienne.fr R?emi Flamary Universit?e C?ote d?Azur, Lagrange, UMR 7293 , CNRS, OCA remi.flamary@unice.fr Nicolas Courty Universit?e de Bretagne Sud, IRISA, U...
6312 |@word mild:1 trial:2 version:8 norm:1 seems:1 nd:1 villani:1 open:3 seek:1 propagate:1 xtest:1 tr:1 initial:1 tuned:1 ours:1 amp:1 comparing:1 nt:4 activation:1 yet:1 must:1 written:1 import:1 readily:1 numerical:1 oberman:1 interpretable:1 half:2 prohibitive:1 realizing:1 short:3 regressive:1 provides:1 recomput...
5,873
6,313
Bayesian Intermittent Demand Forecasting for Large Inventories Matthias Seeger, David Salinas, Valentin Flunkert Amazon Development Center Germany Krausenstrasse 38 10115 Berlin matthis@amazon.de, dsalina@amazon.de, flunkert@amazon.de Abstract We present a scalable and robust Bayesian method for demand forecasting in ...
6313 |@word inversion:1 seems:1 norm:1 nd:1 proportion:1 zit:4 forecaster:2 crucially:1 p0:2 pick:1 harder:1 reduction:4 series:13 contains:3 tuned:2 ours:1 franklin:1 outperforms:5 current:2 comparing:2 z2:1 yet:4 flunkert:2 written:1 john:2 additive:1 happen:1 enables:1 drop:2 treating:1 plot:1 v:1 intelligence:2 ite...
5,874
6,314
Convex Two-Layer Modeling with Latent Structure Vignesh Ganapathiraman? , Xinhua Zhang? , Yaoliang Yu? , Junfeng Wen] ? University of Illinois at Chicago, Chicago, IL, USA ? University of Waterloo, Waterloo, ON, Canada, ] University of Alberta, Edmonton, AB, Canada {vganap2, zhangx}@uic.edu, yaoliang.yu@uwaterloo.ca, ...
6314 |@word version:1 polynomial:2 seems:1 norm:6 yi0:2 open:1 p0:1 contrastive:3 q1:2 sgd:1 tr:11 inpainting:6 accommodate:1 initial:1 score:3 mi0:1 tuned:2 past:1 current:1 com:1 recovered:1 gmail:1 scatter:1 written:1 parsing:2 chicago:2 partition:3 subsequent:2 kdd:1 remove:1 designed:2 drop:1 gist:1 maxv:1 aside:1...
5,875
6,315
Deep Learning Games Dale Schuurmans? Google daes@ualberta.ca Martin Zinkevich Google martinz@google.com Abstract We investigate a reduction of supervised learning to game playing that reveals new connections and learning methods. For convex one-layer problems, we demonstrate an equivalence between global minimizers ...
6315 |@word nd:1 diametrically:1 simplifying:2 sgd:6 accommodate:1 shot:2 recursively:1 reduction:12 uncovered:1 contains:1 selecting:1 series:1 denoting:1 tuned:3 document:1 outperforms:1 current:3 com:1 surprising:2 luo:1 activation:8 si:1 must:3 chicago:2 hajnal:1 gv:5 plot:1 update:5 intelligence:3 warmuth:2 accord...
5,876
6,316
Satisfying Real-world Goals with Dataset Constraints Gabriel Goh Dept. of Mathematics UC Davis Davis, CA 95616 ggoh@math.ucdavis.edu Andrew Cotter, Maya Gupta Google Inc. 1600 Amphitheatre Parkway Mountain View, CA 94043 acotter@google.com mayagupta@google.com Michael Friedlander Dept. of Computer Science University...
6316 |@word version:1 proportion:5 d2:9 seek:3 covariance:1 hsieh:1 sgd:1 liblinear:2 initial:4 series:1 score:1 contains:3 disparity:1 outperforms:1 current:2 com:3 must:3 written:1 fn:3 partition:1 informative:1 enables:1 hypothesize:1 plot:2 maxv:2 v:3 intelligence:1 selected:1 plane:7 mccallum:2 ith:1 math:1 hyperp...
5,877
6,317
?Congruent? and ?Opposite? Neurons: Sisters for Multisensory Integration and Segregation Wen-Hao Zhang1,2 ? , He Wang1 , K. Y. Michael Wong1 , Si Wu2 wenhaoz@ust.hk, hwangaa@connect.ust.hk, phkywong@ust.hk, wusi@bnu.edu.cn 1 Department of Physics, Hong Kong University of Science and Technology, Hong Kong. 2 State Key ...
6317 |@word kong:3 integrative:1 carry:2 disparity:21 idg:1 hereafter:1 interestingly:2 past:2 current:2 recovered:1 comparing:2 si:12 ust:3 written:2 realize:1 ronald:1 informative:2 shape:1 hypothesize:1 medial:5 n0:3 v:5 implying:2 cue:110 half:2 discrimination:1 reciprocal:13 short:1 provides:1 zhang:3 height:2 dir...
5,878
6,318
Synthesis of MCMC and Belief Propagation Sungsoo Ahn? Michael Chertkov? Jinwoo Shin? ? School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea ?1 Theoretical Division, T-4 & Center for Nonlinear Studies, Los Alamos National Laboratory, Los Alamos, NM 87545, USA, ?2 Skolkovo...
6318 |@word mild:1 trial:2 determinant:1 version:3 polynomial:15 calculus:4 decomposition:2 pick:2 harder:1 reduction:1 series:21 outperforms:4 freitas:1 current:3 z2:37 comparing:1 com:1 partition:11 remove:1 plot:1 update:4 stationary:1 intelligence:4 selected:2 advancement:1 core:3 provides:8 iterates:1 completeness...
5,879
6,319
Reshaped Wirtinger Flow for Solving Quadratic System of Equations Huishuai Zhang Department of EECS Syracuse University Syracuse, NY 13244 hzhan23@syr.edu Yingbin Liang Department of EECS Syracuse University Syracuse, NY 13244 yliang06@syr.edu Abstract We study the problem of recovering a vector x ? Rn from its magn...
6319 |@word trial:7 version:2 advantageous:1 norm:2 yi0:1 c0:4 moment:1 shechtman:2 reduction:1 contains:1 initial:2 document:1 ati:31 outperforms:3 recovered:1 z2:2 comparing:1 activation:1 written:1 readily:1 numerical:4 designed:4 plot:1 update:6 intelligence:1 fewer:1 plane:1 xk:5 core:1 fa9550:1 characterization:1...
5,880
632
Analog VLSI Implementation of Multi-dimensional Gradient Descent David B. Kirk, Douglas Kerns, Kurt Fleischer, Alan H. Barr California Institute of Technology Beckman Institute 350-74 Pasadena, CA 91125 E-mail: dkIDegg.gg . cal tech. edu Abstract We describe an analog VLSI implementation of a multi-dimensional gradien...
632 |@word version:1 additively:2 simulation:2 tr:2 initial:1 document:1 kurt:2 yet:1 chu:1 written:1 john:1 realize:1 designed:1 pacemaker:1 dembo:3 filtered:3 provides:4 node:1 along:1 uyi:1 microchip:1 yuhas:1 combine:1 alspector:10 multi:15 integrator:2 chi:2 decomposed:1 decreasing:1 automatically:1 actual:1 provi...
5,881
6,320
Efficient state-space modularization for planning: theory, behavioral and neural signatures Daniel McNamee, Daniel Wolpert, M?t? Lengyel Computational and Biological Learning Lab Department of Engineering University of Cambridge Cambridge CB2 1PZ, United Kingdom {d.mcnamee|wolpert|m.lengyel}@eng.cam.ac.uk Abstract Ev...
6320 |@word version:1 inversion:1 compression:21 seems:1 hippocampus:1 nd:1 termination:3 grey:1 integrative:1 simulation:3 r:1 eng:1 decomposition:2 pg:4 pressure:1 thereby:2 initial:1 series:2 fragment:1 united:1 score:2 daniel:2 mag:2 elaborating:1 task1:1 current:2 contextual:1 od:2 si:6 scatter:1 activation:1 real...
5,882
6,321
RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism Edward Choi? , Mohammad Taha Bahadori? , Joshua A. Kulas? , Andy Schuetz? , Walter F. Stewart? , Jimeng Sun? ? ? Georgia Institute of Technology Sutter Health {mp2893,bahadori,jkulas3}@gatech.edu, {schueta1,stewarwf}@sutterh...
6321 |@word h:1 mild:1 version:1 hu:1 confirms:1 r:1 bn:8 liu:1 series:1 score:1 offering:1 genetic:1 past:6 outperforms:1 existing:1 medi:2 com:2 comparing:1 yet:1 dx:2 must:1 gpu:1 timestamps:3 benign:2 kdd:2 remove:1 drop:1 interpretable:13 v:1 alone:1 stationary:2 selected:3 affair:1 sutter:2 vanishing:1 short:1 re...
5,883
6,322
Exponential expressivity in deep neural networks through transient chaos Ben Poole1 , Subhaneil Lahiri1 , Maithra Raghu2 , Jascha Sohl-Dickstein2 , Surya Ganguli1 1 Stanford University, 2 Google Brain {benpoole,sulahiri,sganguli}@stanford.edu, {maithra,jaschasd}@google.com Abstract We combine Riemannian geometry with...
6322 |@word cox:1 polynomial:2 norm:1 nd:1 open:1 dz1:1 simulation:7 propagate:4 covariance:1 thereby:3 solid:4 moment:1 initial:3 uncovered:1 com:2 wd:1 z2:3 activation:1 bd:1 john:1 numerical:1 partition:1 enables:3 remove:1 update:1 v:1 alone:1 intelligence:1 parameterization:2 plane:6 footing:1 provides:2 pascanu:1...
5,884
6,323
Learnable Visual Markers Oleg Grinchuk1 , Vadim Lebedev1,2 , and Victor Lempitsky1 1 Skolkovo Institute of Science and Technology, Moscow, Russia 2 Yandex, Moscow, Russia Abstract We propose a new approach to designing visual markers (analogous to QR-codes, markers for augmented reality, and robotic fiducial tags) ba...
6323 |@word briefly:1 inversion:1 printer:1 tried:3 bn:1 invoking:1 minus:1 reduction:1 contains:1 series:1 optically:1 tuned:1 interestingly:1 deconvolutional:1 rightmost:1 existing:2 recovered:2 activation:1 synthesizer:29 subsequent:1 blur:7 designed:3 update:1 generative:3 half:1 device:2 core:1 provides:1 location...
5,885
6,324
Local Maxima in the Likelihood of Gaussian Mixture Models: Structural Results and Algorithmic Consequences Chi Jin UC Berkeley chijin@cs.berkeley.edu Yuchen Zhang UC Berkeley yuczhang@berkeley.edu Martin J. Wainwright UC Berkeley wainwrig@berkeley.edu Sivaraman Balakrishnan Carnegie Mellon University siva@stat.cmu....
6324 |@word version:1 polynomial:2 achievable:2 norm:2 suitably:1 open:8 covariance:4 decomposition:2 thereby:1 carry:1 moment:3 initial:5 liu:3 configuration:6 contains:1 selecting:2 series:1 daniel:1 necessity:2 wainwrig:1 current:1 surprising:1 yet:1 written:1 must:3 john:1 transcendental:1 benign:1 plot:2 update:10...
5,886
6,325
Hierarchical Clustering via Spreading Metrics Aurko Roy1 and Sebastian Pokutta2 1 College of Computing, Georgia Institute of Technology, Atlanta, GA, USA. Email: aurko@gatech.edu 2 ISyE, Georgia Institute of Technology, Atlanta, GA, USA. Email: sebastian.pokutta@isye.gatech.edu Abstract We study the cost function for...
6325 |@word repository:1 version:1 polynomial:10 leighton:2 open:2 decomposition:3 citeseer:1 multicommodity:2 recursively:1 contains:1 series:1 lichman:1 daniel:1 existing:1 err:2 assigning:1 must:2 partition:2 drop:1 n0:4 leaf:13 characterization:8 provides:1 node:3 org:1 dn:1 constructed:1 symposium:7 descendant:2 p...
5,887
6,326
Fast and accurate spike sorting of high-channel count probes with KiloSort Marius Pachitariu1 , Nick Steinmetz1 , Shabnam Kadir1 Matteo Carandini1 and Kenneth Harris1 1 UCL, UK {ucgtmpa, }@ucl.ac.uk Abstract New silicon technology is enabling large-scale electrophysiological recordings in vivo from hundreds to thousa...
6326 |@word neurophysiology:2 private:4 achievable:2 norm:3 bf:1 mehta:1 covariance:5 decomposition:6 reduction:1 bai:1 initial:1 score:13 daniel:2 past:2 current:3 com:1 analysed:1 assigning:1 must:2 gpu:3 distant:1 shape:2 remove:2 update:3 v:3 discrimination:1 generative:3 greedy:2 selected:1 fewer:1 intelligence:1 ...
5,888
6,327
Full-Capacity Unitary Recurrent Neural Networks Scott Wisdom1? , Thomas Powers1? , John R. Hershey2 , Jonathan Le Roux2 , and Les Atlas1 1 Department of Electrical Engineering, University of Washington {swisdom, tcpowers, atlas}@uw.edu 2 Mitsubishi Electric Research Laboratories (MERL) {hershey, leroux}@merl.com Abst...
6327 |@word trial:1 timefrequency:1 norm:1 replicate:1 open:1 d2:1 confirms:1 mitsubishi:1 decomposition:1 covariance:1 thereby:1 tr:1 initial:1 series:1 score:1 past:3 imaginary:2 current:1 com:3 z2:2 blank:4 activation:2 must:3 john:1 atlas:2 drop:1 update:3 designed:1 half:2 fewer:1 guess:1 parameterization:16 begin...
5,889
6,328
The Generalized Reparameterization Gradient Francisco J. R. Ruiz University of Cambridge Columbia University Michalis K. Titsias Athens University of Economics and Business David M. Blei Columbia University Abstract The reparameterization gradient has become a widely used method to obtain Monte Carlo gradients to o...
6328 |@word determinant:1 version:1 seems:2 logit:10 simulation:2 covariance:1 moment:3 reduction:3 score:9 document:2 fa8750:1 outperforms:4 com:1 must:1 written:2 john:1 shape:7 analytic:2 christian:1 update:1 aside:1 intelligence:5 generative:1 device:1 inspection:1 short:1 blei:7 provides:4 completeness:1 ire:1 loc...
5,890
6,329
?Short-Dot?: Computing Large Linear Transforms Distributedly Using Coded Short Dot Products Sanghamitra Dutta Carnegie Mellon University sanghamd@andrew.cmu.edu Viveck Cadambe Pennsylvania State University viveck@engr.psu.edu Pulkit Grover Carnegie Mellon University pgrover@andrew.cmu.edu Abstract Faced with satura...
6329 |@word inversion:1 grey:1 km:1 decomposition:2 jacob:1 atrix:1 reduction:1 electronics:1 cyclic:1 liu:1 exclusively:1 necessity:1 outperforms:1 existing:3 com:1 nt:1 si:4 scatter:1 written:1 must:1 nanoscale:1 partition:3 predetermined:1 designed:2 plot:1 sponsored:1 selected:1 device:1 short:54 provides:2 complet...
5,891
633
Planar Hidden Markov Modeling: from Speech to Optical Character Recognition Esther Levin and Roberto Pieraccini AIT Bell Laboratories 600 Mountain Ave. Murray Hill, NJ 07974 Abstract We propose in this paper a statistical model (planar hidden Markov model PHMM) describing statistical properties of images. The model g...
633 |@word trial:1 eliminating:2 polynomial:4 grey:1 tr:2 harder:2 ld:1 initial:1 contains:1 document:1 current:1 assigning:1 yet:1 must:2 written:5 blur:1 drop:1 update:1 selected:1 yr:9 parametrization:1 sudden:1 coarse:1 successive:1 direct:1 become:1 introduce:1 deteriorate:1 behavior:3 decreasing:1 goldman:1 incre...
5,892
6,330
Optimal Architectures in a Solvable Model of Deep Networks Jonathan Kadmon The Racah Institute of Physics and ELSC The Hebrew University, Israel jonathan.kadmon@mail.huji.ac.il Haim Sompolinsky The Racah Institute of Physics and ELSC The Hebrew University, Israel and Center for Brain Science Harvard University Abstra...
6330 |@word neurophysiology:1 version:5 simulation:4 propagate:2 q1:2 thereby:1 solid:1 carry:2 initial:14 series:1 denoting:1 interestingly:1 suppressing:1 kurt:1 past:2 comparing:1 nt:7 mushroom:1 dx:1 numerical:1 subsequent:1 realistic:3 plasticity:1 update:1 n0:11 implying:1 mccallum:1 realism:1 completeness:1 node...
5,893
6,331
Robustness of classifiers: from adversarial to random noise Alhussein Fawzi?, Seyed-Mohsen Moosavi-Dezfooli?, Pascal Frossard ?cole Polytechnique F?d?rale de Lausanne Lausanne, Switzerland {alhussein.fawzi, seyed.moosavi, pascal.frossard} at epfl.ch Abstract Several recent works have shown that state-of-the-art classi...
6331 |@word moosavi:3 dkr:2 version:1 norm:4 seems:1 valle:1 open:2 q1:1 contains:1 mag:2 interestingly:1 bhattacharyya:1 suppressing:1 luo:1 activation:1 intriguing:1 written:2 bd:1 gpu:1 numerical:1 distant:1 remove:1 designed:1 plot:1 plane:2 yamada:1 provides:2 mannor:2 kingsbury:1 along:2 c2:2 mathematical:1 diffe...
5,894
6,332
Geometric Dirichlet Means algorithm for topic inference Mikhail Yurochkin Department of Statistics University of Michigan moonfolk@umich.edu XuanLong Nguyen Department of Statistics University of Michigan xuanlong@umich.edu Abstract We propose a geometric algorithm for topic learning and inference that is built on t...
6332 |@word kulis:4 version:3 proportion:8 seems:1 nd:1 simulation:5 contraction:3 decomposition:1 accounting:1 pick:1 carry:2 moment:1 inefficiency:2 contains:1 liu:2 series:1 tuned:2 document:52 outperforms:4 dx:3 reminiscent:1 must:1 realistic:1 partition:1 hofmann:3 enables:1 aside:1 generative:1 intelligence:1 acc...
5,895
6,333
Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning Mehdi Sajjadi Mehran Javanmardi Tolga Tasdizen Department of Electrical and Computer Engineering University of Utah {mehdi, mehran, tolga}@sci.utah.edu Abstract Effective convolutional neural networks are trained on ...
6333 |@word version:4 norm:1 tried:2 rgb:1 decomposition:1 citeseer:2 pick:1 sajjadi:2 liu:1 contains:9 tuned:1 document:1 com:1 activation:1 must:2 readily:1 cheap:1 v:4 generative:2 fewer:1 intelligence:1 bissacco:1 provides:1 node:2 location:1 simpler:1 zhang:2 five:9 along:2 c2:7 replication:3 consists:4 combine:2 ...
5,896
6,334
Flexible Models for Microclustering with Application to Entity Resolution Giacomo Zanella? Department of Decision Sciences Bocconi University Brenda Betancourt? Department of Statistical Science Duke University giacomo.zanella@unibocconi.it bb222@stat.duke.edu Hanna Wallach Microsoft Research hanna@dirichlet.net ...
6334 |@word proportion:2 sex:1 contains:2 uma:1 series:1 united:1 longitudinal:1 existing:1 err:1 current:1 z2:2 must:1 john:1 realistic:2 partition:31 informative:1 shape:1 pertinent:1 plot:1 generative:1 intelligence:1 parameterization:1 accordingly:1 prize:1 record:38 five:2 mathematical:1 along:1 constructed:2 comm...
5,897
6,335
Deep Alternative Neural Network: Exploring Contexts as Early as Possible for Action Recognition Jinzhuo Wang, Wenmin Wang, Xiongtao Chen, Ronggang Wang, Wen Gao? School of Electronics and Computer Engineering, Peking University ? School of Electronics Engineering and Computer Science, Peking University jzwang@pku.edu....
6335 |@word torsten:2 cnn:7 version:1 wiesel:2 laurence:1 bptt:1 km:1 hu:2 rgb:4 sgd:1 initial:1 configuration:4 contains:4 fragment:4 score:5 liu:1 electronics:2 tuned:2 ours:2 outperforms:2 existing:1 current:6 comparing:1 guadarrama:1 anne:1 parsing:1 ronan:2 christian:1 designed:1 selected:1 amir:1 beginning:1 ith:...
5,898
6,336
Boosting with Abstention Corinna Cortes Google Research New York, NY 10011 Giulia DeSalvo Courant Institute New York, NY 10012 Mehryar Mohri Courant Institute and Google New York, NY 10012 corinna@google.com desalvo@cims.nyu.edu mohri@cims.nyu.edu Abstract We present a new boosting algorithm for the key scenario ...
6336 |@word repository:1 version:2 middle:1 polynomial:1 dubuisson:1 d2:3 incurs:1 thereby:1 configuration:1 series:3 contains:1 document:1 dubourg:1 outperforms:1 com:1 comparing:1 z2:7 luo:1 assigning:1 tackling:1 numerical:2 partition:1 plot:4 designed:1 update:1 selected:2 guess:1 leaf:1 provides:2 boosting:16 iter...
5,899
6,337
Dueling Bandits: Beyond Condorcet Winners to General Tournament Solutions Siddartha Ramamohan Indian Institute of Science Bangalore 560012, India Arun Rajkumar Xerox Research Bangalore 560103, India Shivani Agarwal University of Pennsylvania Philadelphia, PA 19104, USA siddartha.yr@csa.iisc.ernet.in arun_r@csa.iisc...
6337 |@word katja:1 trial:9 exploitation:1 middle:2 open:1 atul:1 decomposition:1 fabrice:1 initial:1 uncovered:18 score:1 contains:2 disparity:1 existing:1 savage:5 contextual:2 current:2 must:3 j1:2 hofmann:1 ramamohan:1 designed:1 update:3 intelligence:1 selected:1 yr:1 instantiate:1 guess:1 website:1 record:1 color...