Unnamed: 0
int64
0
7.24k
id
int64
1
7.28k
raw_text
stringlengths
9
124k
vw_text
stringlengths
12
15k
4,500
5,072
Efficient Online Inference for Bayesian Nonparametric Relational Models Dae Il Kim1 , Prem Gopalan2 , David M. Blei2 , and Erik B. Sudderth1 1 2 Department of Computer Science, Brown University, {daeil,sudderth}@cs.brown.edu Department of Computer Science, Princeton University, {pgopalan,blei}@cs.princeton.edu Abstr...
5072 |@word illustrating:1 briefly:1 open:1 crucially:1 thereby:1 initial:3 configuration:1 contains:2 score:8 zij:3 ka:13 current:1 assigning:1 must:2 realistic:1 subsequent:1 partition:3 informative:2 shape:1 interpretable:1 update:17 v:3 discovering:1 advancement:1 website:1 yamada:1 colored:1 blei:7 provides:4 node...
4,501
5,073
Learning with Noisy Labels Nagarajan Natarajan Inderjit S. Dhillon Pradeep Ravikumar Department of Computer Science, University of Texas, Austin. {naga86,inderjit,pradeepr}@cs.utexas.edu Ambuj Tewari Department of Statistics, University of Michigan, Ann Arbor. tewaria@umich.edu Abstract In this paper, we theoreticall...
5073 |@word trial:1 version:3 norm:1 stronger:1 nd:1 suitably:4 dekel:1 bylander:2 contraction:1 decomposition:1 pavel:1 biconjugate:5 harder:3 reduction:1 liu:4 lucet:2 tuned:1 interestingly:1 existing:1 analysed:1 yet:1 liva:1 realize:1 numerical:1 benign:1 shape:1 plot:5 juditsky:1 aside:1 instantiate:1 accordingly:...
4,502
5,074
Low-rank matrix reconstruction and clustering via approximate message passing Ryosuke Matsushita NTT DATA Mathematical Systems Inc. 1F Shinanomachi Rengakan, 35, Shinanomachi, Shinjuku-ku, Tokyo, 160-0016, Japan matsur8@gmail.com Toshiyuki Tanaka Department of Systems Science, Graduate School of Informatics, Kyoto Un...
5074 |@word trial:9 norm:1 nd:1 km:9 bvt:3 covariance:2 thereby:1 tr:1 initial:7 liu:1 contains:1 document:1 amp:53 outperforms:2 existing:1 com:1 si:4 gmail:1 additive:1 numerical:5 kdd:1 analytic:1 seeding:1 update:4 maxv:1 aside:1 intelligence:1 nearness:1 toronto:1 five:3 mathematical:1 symposium:4 incorrect:1 cons...
4,503
5,075
Sign Cauchy Projections and Chi-Square Kernel Ping Li Dept of Statistics & Biostat. Dept of Computer Science Rutgers University pingli@stat.rutgers.edu Gennady Samorodnitsky ORIE and Dept of Stat. Science Cornell University Ithaca, NY 14853 gs18@cornell.edu John Hopcroft Dept of Computer Science Cornell University It...
5075 |@word multitask:1 norm:2 open:1 confirms:1 simulation:2 hsieh:1 gennady:3 sgd:1 euclidian:1 solid:5 reduction:2 liblinear:6 hoiem:1 document:1 interestingly:5 bc:1 err:1 com:1 written:1 john:2 concatenate:3 additive:1 confirming:1 kdd:2 designed:1 drop:1 update:1 plot:4 hash:3 short:2 fa9550:1 provides:5 quantize...
4,504
5,076
Learning Multi-level Sparse Representations Ferran Diego Fred A. Hamprecht Heidelberg Collaboratory for Image Processing (HCI) Interdisciplinary Center for Scientific Computing (IWR) University of Heidelberg, Heidelberg 69115, Germany {ferran.diego,fred.hamprecht}@iwr.uni-heidelberg.de Abstract Bilinear approximation...
5076 |@word middle:1 eliminating:1 achievable:1 seems:3 norm:8 plsa:1 tensorial:1 simulation:1 decomposition:16 q1:12 schnitzer:1 moment:1 electronics:1 selecting:1 document:1 rightmost:2 activation:10 intriguing:1 written:1 must:1 subsequent:1 shape:4 designed:2 plot:2 greedy:2 pursued:1 intelligence:1 imitate:1 accor...
4,505
5,077
A New Convex Relaxation for Tensor Completion Bernardino Romera-Paredes Department of Computer Science and UCL Interactive Centre University College London Malet Place, London WC1E 6BT, UK B.RomeraParedes@cs.ucl.ac.uk Massimiliano Pontil Department of Computer Science and Centre for Computational Statistics and Machi...
5077 |@word multitask:2 compression:1 advantageous:1 norm:48 paredes:3 open:1 calculus:1 tried:3 bn:21 decomposition:6 rgb:1 thereby:1 initial:2 liu:1 contains:1 romera:3 current:1 wd:1 com:1 si:2 chu:1 john:1 numerical:4 visible:1 j1:2 designed:1 update:1 rpn:1 selected:1 rp1:10 paulin:1 yamada:1 provides:2 authority:...
4,506
5,078
Latent Maximum Margin Clustering Guang-Tong Zhou, Tian Lan, Arash Vahdat, and Greg Mori School of Computing Science Simon Fraser University {gza11,tla58,avahdat,mori}@cs.sfu.ca Abstract We present a maximum margin framework that clusters data using latent variables. Using latent representations enables our framework ...
5078 |@word briefly:2 km:6 tried:1 arti:1 initial:1 liu:2 wedding:7 efficacy:4 score:18 contains:2 denoting:1 outperforms:4 current:4 yet:1 assigning:1 written:1 partition:1 hofmann:1 enables:2 remove:1 update:1 grass:2 intelligence:2 instantiate:3 website:1 selected:2 discovering:1 plane:2 short:2 detecting:2 provides...
4,507
5,079
Statistical analysis of coupled time series with Kernel Cross-Spectral Density operators. Michel Besserve MPI for Intelligent Systems and MPI for Biological Cybernetics, T?ubingen, Germany michel.besserve@tuebingen.mpg.de Nikos K. Logothetis MPI for Biological Cybernetics, T?ubingen nikos.logothetis@tuebingen.mpg.de ...
5079 |@word h:9 mild:2 trial:2 version:1 middle:3 norm:21 proportion:1 hyv:1 simulation:1 accounting:1 covariance:7 decomposition:1 pick:1 tr:7 solid:1 reduction:1 tapering:3 initial:1 series:63 contains:2 rkhs:6 interestingly:2 current:1 si:1 subsequent:1 enables:4 reproducible:1 plot:1 stationary:4 intelligence:1 sel...
4,508
508
A Topographic Product for the Optimization of Self-Organizing Feature Maps Hans-Ulrich Bauer, Klaus Pawelzik, Theo Geisel Institut fUr theoretische Physik and SFB Nichtlineare Dynamik Universitat Frankfurt Robert-Mayer-Str. 8-10 W -6000 Frankfurt 11 Fed. Rep . of Germany email: bauer@asgard.physik.uni-frankfurt.dbp A...
508 |@word briefly:1 seems:1 physik:2 pick:1 nt:2 must:1 subsequent:1 speakerindependent:1 plot:1 v:1 nichtlineare:2 node:9 brandt:2 introduce:1 roughly:1 brain:1 pawelzik:6 str:1 becomes:1 underlying:1 notation:1 suffice:1 deutsche:1 what:1 dynamik:2 minimizes:1 interpreted:1 q2:2 quantitative:1 demonstrates:1 classif...
4,509
5,080
Robust Low Rank Kernel Embeddings of Multivariate Distributions Le Song, Bo Dai College of Computing, Georgia Institute of Technology lsong@cc.gatech.edu, bodai@gatech.edu Abstract Kernel embedding of distributions has led to many recent advances in machine learning. However, latent and low rank structures prevalent i...
5080 |@word repository:1 norm:5 nd:2 ci2:2 decomposition:42 covariance:2 pick:1 dramatic:1 recursively:2 ld:1 carry:3 moment:1 contains:1 rkhs:5 xnj:1 current:1 z2:14 si:3 yet:1 dx:2 written:1 readily:3 bd:1 subsequent:1 concatenate:1 partition:2 numerical:1 shape:1 designed:1 plot:1 bickson:1 v:1 intelligence:1 leaf:1...
4,510
5,081
B-tests: Low Variance Kernel Two-Sample Tests Matthew Blaschko Arthur Gretton Wojciech Zaremba ? Gatsby Unit Center for Visual Computing Equipe GALEN ? University College London Inria Saclay Ecole Centrale Paris United Kingdom Ch?atenay-Malabry, France Ch?atenay-Malabry, France {woj.zaremba,arthur.gretton}@gmail.com, ...
5081 |@word briefly:1 mmds:1 smirnov:2 nd:1 open:1 bn:3 covariance:5 harder:1 moment:5 series:2 united:1 ecole:1 rkhs:5 bootstrapped:1 existing:1 com:2 exy:1 gmail:1 yet:2 must:1 written:2 fn:2 visible:3 oldenbourg:1 plot:1 drop:2 v:3 half:2 selected:1 fewer:1 es:1 ith:1 core:1 yamada:1 accepting:1 location:1 simpler:2...
4,511
5,082
On Flat versus Hierarchical Classification in Large-Scale Taxonomies Rohit Babbar, Ioannis Partalas, Eric Gaussier, Massih-Reza Amini Universit? Joseph Fourier, Laboratoire Informatique de Grenoble BP 53 - F-38041 Grenoble Cedex 9 firstname.lastname@imag.fr Abstract We study in this paper flat and hierarchical classi...
5082 |@word version:9 norm:1 nd:1 dekel:2 hsieh:1 bioasq:1 tr:1 harder:1 liblinear:2 bai:1 liu:3 series:2 denoting:2 document:5 academia:1 hofmann:1 designed:1 rd2:2 v:2 leaf:2 selected:2 directory:1 gfb:5 record:1 provides:3 node:41 org:1 zhang:1 along:1 direct:2 descendant:1 consists:1 introduce:2 indeed:2 behavior:1...
4,512
5,083
Robust Bloom Filters for Large Multilabel Classification Tasks Moustapha Ciss?e LIP6, UPMC Sorbonne Universit?e Paris, France first.last@lip6.fr Nicolas Usunier UT Compi`egne, CNRS Heudiasyc UMR 7253 Compi`egne, France nusunier@utc.fr Thierry Artieres, Patrick Gallinari LIP6, UPMC Sorbonne Universit?e Paris, France f...
5083 |@word version:1 middle:3 compression:3 stronger:1 nd:1 bf:29 dekel:6 hsieh:1 p0:1 recursively:1 liblinear:2 reduction:9 configuration:1 contains:1 score:5 hereafter:1 document:1 ours:1 bitwise:3 existing:1 com:1 must:1 partition:3 confirming:2 cis:1 remove:1 designed:4 v:2 hash:23 half:1 leaf:2 selected:1 egne:2 ...
4,513
5,084
Gaussian Process Conditional Copulas with Applications to Financial Time Series James Robert Lloyd Engineering Department University of Cambridge jrl44@cam.ac.uk Jos?e Miguel Hern?andez-Lobato Engineering Department University of Cambridge jmh233@cam.ac.uk Daniel Hern?andez-Lobato Computer Science Department Univers...
5084 |@word version:2 middle:4 csx:1 covariance:11 decomposition:1 accommodate:2 reduction:2 moment:5 configuration:1 series:29 initial:1 daniel:3 outperforms:1 current:1 parameter1:1 cad:2 si:2 worsening:1 written:1 numerical:4 subsequent:1 plot:8 drop:1 update:4 n0:6 statis:1 alone:1 intelligence:3 prohibitive:1 para...
4,514
5,085
Bayesian Inference and Learning in Gaussian Process State-Space Models with Particle MCMC Roger Frigola1 , Fredrik Lindsten2 , Thomas B. Sch?on2,3 and Carl E. Rasmussen1 1. Dept. of Engineering, University of Cambridge, UK, {rf342,cer54}@cam.ac.uk 2. Div. of Automatic Control, Link?oping University, Sweden, lindsten@is...
5085 |@word briefly:1 version:1 proportionality:1 seek:1 propagate:1 crucially:2 covariance:10 simulation:4 fifteen:1 harder:1 moment:1 reduction:1 liu:2 series:3 contains:1 interestingly:1 existing:1 ka:1 readily:1 subsequent:1 drop:1 plot:1 update:4 resampling:2 v:1 generative:1 selected:1 greedy:1 prohibitive:1 inte...
4,515
5,086
Multi-Task Bayesian Optimization Kevin Swersky Department of Computer Science University of Toronto kswersky@cs.toronto.edu Jasper Snoek? School of Engineering and Applied Sciences Harvard University jsnoek@seas.harvard.edu Ryan P. Adams School of Engineering and Applied Sciences Harvard University rpa@seas.harvard....
5086 |@word exploitation:2 faculty:1 version:4 cnn:4 cox:1 retraining:1 mockus:1 willing:1 zilinskas:1 rgb:1 covariance:5 pick:3 solid:2 reduction:2 initial:1 configuration:1 contains:1 score:2 selecting:2 daniel:1 bootstrapped:2 document:2 existing:1 freitas:2 contextual:2 com:2 must:1 john:1 fn:2 cheap:2 christian:1 ...
4,516
5,087
Efficient Optimization for Sparse Gaussian Process Regression Yanshuai Cao1 1 Marcus A. Brubaker2 David J. Fleet1 2 Department of Computer Science University of Toronto TTI-Chicago Aaron Hertzmann1,3 3 Adobe Research Abstract We propose an efficient optimization algorithm for selecting a subset of training data...
5087 |@word version:1 inversion:2 dalal:1 norm:2 triggs:1 termination:2 covariance:9 decomposition:4 nystr:4 tr:6 harder:1 reduction:7 initial:1 score:4 selecting:4 disparity:1 outperforms:3 existing:1 current:3 comparing:1 chu:1 chicago:1 informative:2 remove:3 update:15 v:1 greedy:6 prohibitive:2 selected:1 ith:1 pro...
4,517
5,088
Variational Inference for Mahalanobis Distance Metrics in Gaussian Process Regression Michalis K. Titsias Department of Informatics Athens University of Economics and Business mtitsias@aueb.gr Miguel L?azaro-Gredilla Dpt. Signal Processing & Communications Universidad Carlos III de Madrid - Spain miguel@tsc.uc3m.es ...
5088 |@word determinant:1 version:3 middle:1 inversion:2 seems:1 simulation:1 covariance:5 solid:1 reduction:8 initial:2 contains:1 initialisation:1 tuned:1 existing:3 current:1 wd:1 written:2 must:1 realistic:1 wx:12 remove:1 v:1 intelligence:1 selected:1 guess:1 discovering:1 maximised:1 provides:2 node:2 toronto:1 w...
4,518
5,089
It is all in the noise: Efficient multi-task Gaussian process inference with structured residuals Christoph Lippert Microsoft Research Los Angeles, USA lippert@microsoft.com Barbara Rakitsch Machine Learning and Computational Biology Research Group Max Planck Institutes T?ubingen, Germany rakitsch@tuebingen.mpg.de Ka...
5089 |@word trial:1 cu:1 simulation:7 covariance:72 decomposition:2 yuc:1 thereby:2 series:1 score:1 genetic:1 outperforms:1 existing:2 com:1 nt:3 written:2 john:1 fn:2 multioutput:1 n0:1 generative:1 selected:4 xk:1 prize:1 core:1 smith:1 filtered:1 math:1 preference:1 simpler:1 zhang:1 five:1 ethanol:3 fitting:2 fals...
4,519
509
Towards Faster Stochastic Gradient Search Christian Darken and John Moody Yale Computer Science, P.O. Box 2158, New Haven, CT 06520 Email: darken@cs.yale.edu Abstract Stochastic gradient descent is a general algorithm which includes LMS, on-line backpropagation, and adaptive k-means clustering as special cases. The s...
509 |@word illustrating:1 jacob:4 paid:1 dramatic:2 initial:1 loc:1 seriously:1 current:1 elliptical:1 od:1 john:1 numerical:1 christian:1 pertinent:1 update:1 v:1 stationary:1 provides:1 math:2 location:2 mathematical:1 direct:1 persistent:1 qualitative:1 prove:1 theoretically:1 roughly:1 behavior:4 nor:1 automaticall...
4,520
5,090
Spike train entropy-rate estimation using hierarchical Dirichlet process priors Karin Knudson Department of Mathematics kknudson@math.utexas.edu Jonathan W. Pillow Center for Perceptual Systems Departments of Psychology & Neuroscience The University of Texas at Austin pillow@mail.utexas.edu Abstract Entropy rate qua...
5090 |@word neurophysiology:1 briefly:1 proportion:1 cs0:1 simulation:1 jacob:1 edric:1 recursively:1 series:4 contains:1 past:1 existing:4 contextual:3 yet:2 must:1 parsing:1 subsequent:1 partition:2 distant:4 discernible:1 plot:3 drop:1 stationary:4 cue:1 xk:3 short:7 memoizer:1 blei:1 provides:2 math:1 node:3 revisi...
4,521
5,091
Designed Measurements for Vector Count Data 1 Liming Wang, 1 David Carlson, 2 Miguel Dias Rodrigues, 3 David Wilcox, 1 Robert Calderbank and 1 Lawrence Carin 1 Department of Electrical and Computer Engineering, Duke University 2 Department of Electronic and Electrical Engineering, University College London 3 Departme...
5091 |@word version:1 compression:1 logit:1 nd:1 c0:1 seek:1 tried:1 pulse:2 simulation:2 r:1 prasad:1 dramatic:2 document:26 mmse:7 existing:2 recovered:1 current:11 od:1 wd:2 readily:1 cruz:1 informative:2 analytic:1 shamai:3 designed:15 update:1 v:1 half:1 selected:2 harmany:1 sys:6 ith:2 blei:1 math:1 org:1 five:1 ...
4,522
5,092
Dirty Statistical Models Eunho Yang Department of Computer Science University of Texas at Austin eunho@cs.utexas.edu Pradeep Ravikumar Department of Computer Science University of Texas at Austin pradeepr@cs.utexas.edu Abstract We provide a unified framework for the high-dimensional analysis of ?superposition-structu...
5092 |@word briefly:1 loading:2 norm:20 stronger:3 decomposition:6 covariance:7 score:3 denoting:1 past:1 ka:1 yet:1 written:2 additive:1 sys:1 caveat:2 provides:1 allerton:1 simpler:2 zhang:1 c2:5 become:1 shorthand:1 theoretically:1 inter:1 expected:1 indeed:2 cand:3 nor:1 p1:2 multi:3 decreasing:1 spherical:1 encour...
4,523
5,093
Summary Statistics for Partitionings and Feature Allocations Is??k Bar?s? Fidaner Computer Engineering Department Bo?gazic?i University, Istanbul fidaner@alternatifbilisim.org Ali Taylan Cemgil Computer Engineering Department Bo?gazic?i University, Istanbul taylan.cemgil@boun.edu.tr Abstract Infinite mixture models ...
5093 |@word version:1 middle:1 proportion:2 open:1 cyprus:1 seek:1 eng:1 innermost:1 pg:4 tr:1 initial:2 liu:1 contains:3 score:1 qatar:1 genetic:1 africa:1 existing:1 current:1 comparing:2 cumulation:2 nt:4 si:6 written:4 indonesia:1 partition:8 informative:3 plot:5 designed:1 update:1 malaysia:1 leaf:3 chile:1 ith:1 ...
4,524
5,094
Dynamic Clustering via Asymptotics of the Dependent Dirichlet Process Mixture Trevor Campbell MIT Cambridge, MA 02139 Miao Liu Duke University Durham, NC 27708 tdjc@mit.edu miao.liu@duke.edu Brian Kulis Ohio State University Columbus, OH 43210 Jonathan P. How MIT Cambridge, MA 02139 Lawrence Carin Duke Universit...
5094 |@word aircraft:8 kulis:3 trial:3 nd:1 vogt:1 d2:1 vermaak:1 recursively:1 reduction:2 liu:2 united:2 selecting:1 tuned:1 longitudinal:1 past:3 outperforms:1 current:11 yet:1 hou:1 john:2 enables:1 analytic:1 jfk:1 christian:2 update:13 generative:2 half:2 selected:1 nq:7 discovering:2 plane:5 scotland:1 wolfram:2...
4,525
5,095
k-Prototype Learning for 3D Rigid Structures ? Hu Ding Department of Computer Science and Engineering State University of New York at Buffalo Buffalo, NY14260 huding@buffalo.edu Ronald Berezney Department of Biological Sciences State University of New York at Buffalo Buffalo, NY14260 berezney@buffalo.edu Jinhui Xu D...
5095 |@word version:2 seems:2 norm:2 vi1:1 hu:1 q1:13 recursively:2 reduction:2 initial:2 configuration:2 contains:4 selecting:1 genetic:1 kahl:1 existing:5 current:2 comparing:2 si:3 must:1 ronald:1 partition:12 shape:10 opin:1 plot:2 update:2 medial:1 selected:2 item:1 realizing:2 core:2 math:1 node:1 firstly:2 c22:7...
4,526
5,096
Distributed k-Means and k-Median Clustering on General Topologies Maria Florina Balcan, Steven Ehrlich, Yingyu Liang School of Computer Science Georgia Institute of Technology Atlanta, GA 30332 {ninamf,sehrlich}@cc.gatech.edu,yliang39@gatech.edu Abstract This paper provides new algorithms for distributed clustering f...
5096 |@word briefly:1 compression:1 polynomial:1 widom:1 heiser:1 pick:1 incurs:1 venkatasubramanian:1 liu:1 outperforms:3 existing:2 surprising:1 si:6 must:1 readily:1 additive:1 partition:11 designed:1 half:1 selected:1 xk:1 fa9550:1 provides:1 node:28 location:1 gx:3 c6:1 zhang:7 five:1 height:5 c2:3 direct:1 become...
4,527
5,097
Multiclass Total Variation Clustering Thomas Laurent Loyola Marymount University Los Angeles, CA 90045 tlaurent@lmu.edu Xavier Bresson University of Lausanne Lausanne, Switzerland xavier.bresson@unil.ch James H. von Brecht University of California, Los Angeles Los Angeles, CA 90095 jub@math.ucla.edu David Uminsky U...
5097 |@word trial:3 dkr:4 version:8 norm:2 seems:1 simulation:1 propagate:1 bn:1 decomposition:1 unimpressive:1 asks:1 tice:1 initial:5 contains:3 series:1 outperforms:1 current:3 yet:1 must:3 readily:1 numerical:2 partition:10 plot:1 n0:1 half:1 selected:1 spec:1 intelligence:2 desktop:1 beginning:1 ith:2 fa9550:1 lr:...
4,528
5,098
Learning Multiple Models via Regularized Weighting Daniel Vainsencher Department of Electrical Engineering Technion, Haifa, Israel danielv@tx.technion.ac.il Shie Mannor Department of Electrical Engineering Technion, Haifa, Israel shie@ee.technion.ac.il Huan Xu Mechanical Engineering Department National University of...
5098 |@word briefly:1 polynomial:1 norm:1 seems:2 proportion:1 reused:1 dekker:1 seek:5 bn:2 elisseeff:1 initial:1 mpexuh:1 series:2 daniel:1 gagliardi:1 current:2 aberrant:1 yet:1 assigning:1 must:7 john:3 additive:1 partition:1 visible:1 happen:1 cheap:1 generative:1 leaf:1 fewer:2 half:1 intelligence:1 plane:1 ith:2...
4,529
5,099
Regularized Spectral Clustering under the Degree-Corrected Stochastic Blockmodel Karl Rohe Department of Statistics University of Wisconsin-Madison Madison, WI karlrohe@stat.wisc.edu Tai Qin Department of Statistics University of Wisconsin-Madison Madison, WI qin@stat.wisc.edu Abstract Spectral clustering is a fast ...
5099 |@word version:7 pw:1 norm:3 proportion:1 nd:1 c0:2 simulation:7 tried:1 decomposition:5 zbl:2 contains:3 score:16 current:2 written:1 john:1 numerical:1 partition:12 informative:1 shape:3 analytic:1 remove:2 plot:1 v:1 xk:2 ith:1 jiashun:1 provides:5 node:71 ames:2 liberal:1 five:1 c2:2 beta:1 symposium:2 natalie...
4,530
51
495 REFLEXIVE ASSOCIATIVE MEMORIES Hendrlcus G. Loos Laguna Research Laboratory, Fallbrook, CA 92028-9765 ABSTRACT In the synchronous discrete model, the average memory capacity of bidirectional associative memories (BAMs) is compared with that of Hopfield memories, by means of a calculat10n of the percentage of good ...
51 |@word norm:3 cha:1 cml:3 heteroassociative:1 accounting:1 thres:2 reduction:1 configuration:1 contains:1 must:2 remove:1 bart:1 half:8 selected:1 device:2 pointer:1 ik:1 soffer:1 combine:1 manner:2 hresholding:1 mechanic:1 project:1 notation:1 mass:1 transformation:6 hypothetical:1 act:2 bipolar:3 um:1 unit:1 appea...
4,531
510
Threshold Network Learning in the Presence of Equivalences John Shawe-Taylor Department of Computer Science Royal Holloway and Bedford New College University of London Egham, Surrey TW20 OEX, UK Abstract This paper applies the theory of Probably Approximately Correct (PAC) learning to multiple output feedforward thre...
510 |@word km:1 simplifying:1 tr:1 substitution:2 selecting:1 chervonenkis:5 lang:1 john:6 fn:1 shawetaylor:1 warmuth:1 manfred:1 slh:1 completeness:1 bijection:1 node:27 prove:2 introduce:2 indeed:1 multi:1 automatically:1 equipped:1 considering:3 begin:2 bounded:2 notation:1 what:1 unified:1 guarantee:1 certainty:1 e...
4,532
5,100
Moment-based Uniform Deviation Bounds for k-means and Friends Matus Telgarsky Sanjoy Dasgupta Computer Science and Engineering, UC San Diego {mtelgars,dasgupta}@cs.ucsd.edu Abstract Suppose k centers are fit to m points by heuristically minimizing the k-means cost; what is the corresponding fit over the source distrib...
5100 |@word manageable:1 polynomial:1 norm:12 duda:1 bf:6 heuristically:1 d2:1 heretofore:1 covariance:8 p0:17 pick:1 mention:1 boundedness:6 carry:1 moment:28 contains:1 score:5 series:1 bc:3 com:1 must:2 written:1 readily:1 subsequent:1 zeger:2 numerical:1 drop:1 update:2 selected:1 guess:2 steepest:1 core:1 farther:...
4,533
5,101
Statistical Active Learning Algorithms Vitaly Feldman IBM Research - Almaden vitaly@post.harvard.edu Maria Florina Balcan Georgia Institute of Technology ninamf@cc.gatech.edu Abstract We describe a framework for designing efficient active learning algorithms that are tolerant to random classification noise and differ...
5101 |@word private:16 faculty:1 version:3 polynomial:7 suitably:1 dekel:1 d2:2 simulation:8 covariance:2 invoking:1 asks:1 harder:1 reduction:2 chervonenkis:1 ours:1 past:1 current:2 beygelzimer:3 chu:1 attracted:1 must:1 informative:1 guess:3 warmuth:1 isotropic:8 mccallum:1 smith:1 core:1 record:4 fa9550:1 filtered:...
4,534
5,102
Predictive PAC Learning and Process Decompositions Aryeh Kontorovich Computer Science Department Ben Gurion University Beer Sheva 84105 Israel karyeh@cs.bgu.ac.il Cosma Rohilla Shalizi Statistics Department Carnegie Mellon University Pittsburgh, PA 15213 USA cshalizi@cmu.edu Abstract We informally call a stochastic ...
5102 |@word version:2 open:2 seek:1 decomposition:7 attainable:1 pick:1 asks:1 initial:3 contains:1 series:2 pub:1 chervonenkis:2 ecole:1 past:14 existing:1 bradley:1 lang:1 written:1 must:12 john:1 fn:1 realistic:1 happen:1 gurion:1 enables:1 drop:1 aside:1 implying:1 stationary:14 selected:2 device:1 item:1 warmuth:1...
4,535
5,103
Adaptivity to Local Smoothness and Dimension in Kernel Regression Samory Kpotufe Toyota Technological Institute-Chicago? samory@ttic.edu Vikas K Garg Toyota Technological Institute-Chicago vkg@ttic.edu Abstract We present the first result for kernel regression where the procedure adapts locally at a point x to both ...
5103 |@word version:1 c0:19 d2:1 simulation:1 decomposition:1 invoking:1 pick:2 selecting:3 tuned:2 existing:1 comparing:1 must:6 fn:33 subsequent:1 chicago:2 designed:1 selected:5 math:1 mcdiarmid:2 mathematical:1 c2:6 ouput:1 prove:2 consists:2 combine:2 introduce:1 x0:5 expected:1 inspired:1 globally:1 decreasing:2 ...
4,536
5,104
A Comparative Framework for Preconditioned Lasso Algorithms Michael I. Jordan Nebojsa Jojic Fabian L. Wauthier Computer Science Division Microsoft Research, Redmond Statistics and WTCHG jojic@microsoft.com University of California, Berkeley University of Oxford jordan@cs.berkeley.edu flw@stats.ox.ac.uk Abstract The L...
5104 |@word briefly:2 eliminating:1 seems:2 norm:2 underline:1 suitably:2 crucially:1 initial:1 series:2 selecting:1 elliptical:1 com:1 comparing:9 recovered:1 yet:1 must:1 readily:2 wanted:1 plot:2 v:16 nebojsa:1 generative:10 instantiate:1 accordingly:1 core:2 underestimating:1 mitigation:1 constructed:2 direct:2 ik:...
4,537
5,105
New Subsampling Algorithms for Fast Least Squares Regression Paramveer S. Dhillon1 Yichao Lu2 Dean Foster2 Lyle Ungar1 1 2 Computer & Information Science, Statistics (Wharton School) University of Pennsylvania, Philadelphia, PA, U.S.A {dhillon|ungar}@cis.upenn.edu foster@wharton.upenn.edu, yichaolu@sas.upenn.edu Abst...
5105 |@word version:1 norm:3 proportion:1 bf:11 covariance:14 decomposition:1 mention:1 solid:3 initial:1 contains:1 comparing:1 chazelle:1 must:1 john:2 plot:2 oldest:1 xk:1 lr:1 provides:1 completeness:1 firstly:1 zhang:1 mathematical:2 constructed:1 c2:4 fitting:2 theoretically:1 upenn:3 expected:1 growing:1 multi:1...
4,538
5,106
Faster Ridge Regression via the Subsampled Randomized Hadamard Transform Yichao Lu1 Paramveer S. Dhillon2 Dean Foster1 Lyle Ungar2 1 2 Statistics (Wharton School), Computer & Information Science University of Pennsylvania, Philadelphia, PA, U.S.A {dhillon|ungar}@cis.upenn.edu foster@wharton.upenn.edu, yichaolu@sas.upe...
5106 |@word trial:1 version:1 inversion:1 norm:3 tried:4 covariance:3 decomposition:1 tr:4 solid:2 recursively:1 reduction:1 contains:1 series:1 selecting:1 daniel:1 current:1 chazelle:1 concatenate:1 subsequent:3 hypothesize:1 v:1 fewer:1 nq:1 caveat:1 firstly:5 zhang:1 constructed:1 ik:4 prove:1 consists:1 introduce:...
4,539
5,107
Sequential Transfer in Multi-armed Bandit with Finite Set of Models Mohammad Gheshlaghi Azar ? Alessandro Lazaric ? School of Computer Science INRIA Lille - Nord Europe CMU Team SequeL Emma Brunskill ? School of Computer Science CMU Abstract Learning from prior tasks and transferring that experience to improve future...
5107 |@word mild:1 trial:1 multitask:2 version:1 norm:4 nd:4 dekel:1 open:2 simulation:3 decomposition:4 pick:1 incurs:2 reduction:1 moment:13 liu:1 series:2 contains:3 venkatasubramanian:1 outperforms:1 current:7 com:1 recovered:1 contextual:1 yet:2 written:1 numerical:4 j1:1 confirming:1 remove:1 update:3 v:1 station...
4,540
5,108
Prior-free and prior-dependent regret bounds for Thompson Sampling S?ebastien Bubeck, Che-Yu Liu Department of Operations Research and Financial Engineering, Princeton University sbubeck@princeton.edu, cheliu@princeton.edu Abstract We consider the stochastic multi-armed bandit problem with a prior distribution on the...
5108 |@word cu:2 seems:1 nd:1 open:1 simulation:1 decomposition:1 liu:1 series:1 tuned:1 past:2 outperforms:1 current:1 dx:3 must:1 reminiscent:1 numerical:3 realistic:1 remove:1 drop:2 beginning:1 completeness:1 c2:2 beta:3 prove:2 indeed:2 surge:1 multi:7 inspired:4 armed:11 bounded:8 bonus:1 what:2 developed:1 findi...
4,541
5,109
Two-Target Algorithms for Infinite-Armed Bandits with Bernoulli Rewards Thomas Bonald? Department of Networking and Computer Science Telecom ParisTech Paris, France thomas.bonald@telecom-paristech.fr Alexandre Prouti`ere?? Automatic Control Department KTH Stockholm, Sweden alepro@kth.se Abstract We consider an infini...
5109 |@word exploitation:3 version:6 simulation:1 initial:1 contains:4 selecting:1 current:3 nt:1 rocquencourt:1 numerical:2 remove:1 selected:8 metrika:1 item:1 short:3 provides:1 successive:2 preference:1 teytaud:2 mathematical:1 beta:7 incorrect:2 prove:3 consists:3 introduce:1 expected:7 themselves:1 planning:1 mul...
4,542
511
Neural Network Analysis of Event Related Potentials and Electroencephalogram Predicts Vigilance Rita Venturini William W. Lytton Terrence J. Sejnowski Computational Neurobiology Laboratory The Salk Institute La J oBa, CA 92037 Abstract Automated monitoring of vigilance in attention intensive tasks such as air traff...
511 |@word neurophysiology:2 trial:2 determinant:1 instruction:1 pulse:1 simulation:6 solid:2 initial:2 united:1 past:1 motor:1 progressively:1 half:1 device:1 tone:2 short:2 filtered:1 five:2 along:1 profound:1 sustained:2 inter:2 rapid:1 simulator:1 brain:3 decreasing:1 little:1 window:3 electroencephalography:3 prov...
4,543
5,110
Thompson Sampling for 1-Dimensional Exponential Family Bandits Emilie Kaufmann Institut Mines-Telecom; Telecom ParisTech kaufmann@telecom-paristech.fr Nathaniel Korda INRIA Lille - Nord Europe, Team SequeL nathaniel.korda@inria.fr Remi Munos INRIA Lille - Nord Europe, Team SequeL remi.munos@inria.fr Abstract Thomps...
5110 |@word exploitation:1 c0:1 open:1 decomposition:3 p0:3 concise:1 liu:1 contains:1 renewed:1 pna:1 contextual:2 nt:1 varx:1 yet:1 informative:1 shape:1 drop:1 intelligence:1 xk:1 beginning:1 short:1 provides:1 completeness:1 honda:1 along:1 c2:21 direct:1 beta:3 introduce:6 notably:1 indeed:2 expected:3 multi:3 dec...
4,544
5,111
Bayesian Mixture Modeling and Inference based Thompson Sampling in Monte-Carlo Tree Search Aijun Bai Univ. of Sci. & Tech. of China baj@mail.ustc.edu.cn Feng Wu University of Southampton fw6e11@ecs.soton.ac.uk Xiaoping Chen Univ. of Sci. & Tech. of China xpchen@ustc.edu.cn Abstract Monte-Carlo tree search (MCTS) ha...
5111 |@word h:3 exploitation:3 version:1 briefly:3 middle:1 seems:1 hector:1 open:1 simulation:13 recursively:3 initial:2 bai:1 selecting:6 past:3 subjective:1 existing:1 current:6 comparing:4 com:1 dx:1 must:3 guez:1 john:1 realistic:2 informative:2 confirming:1 shlomo:1 update:4 v:1 stationary:5 generative:1 selected...
4,545
5,112
Density estimation from unweighted k-nearest neighbor graphs: a roadmap Ulrike von Luxburg and Morteza Alamgir Department of Computer Science University of Hamburg, Germany {luxburg,alamgir}@informatik.uni-hamburg.de Abstract Consider an unweighted k-nearest neighbor graph on n points that have been sampled i.i.d. fr...
5112 |@word middle:1 version:1 suitably:1 open:1 grey:2 simulation:3 p0:9 harder:1 carry:1 reduction:2 contains:1 score:2 series:1 current:2 si:1 yet:4 dx:2 written:1 drop:1 plot:6 alone:4 half:1 leaf:1 selected:1 short:1 location:1 readability:1 warmup:1 mathematical:2 along:12 constructed:1 direct:1 predecessor:1 bor...
4,546
5,113
Sketching Structured Matrices for Faster Nonlinear Regression David P. Woodruff IBM Almaden Research Center San Jose, CA 95120 dpwoodru@us.ibm.com Haim Avron Vikas Sindhwani IBM T.J. Watson Research Center Yorktown Heights, NY 10598 {haimav,vsindhw}@us.ibm.com Abstract Motivated by the desire to extend fast randomiz...
5113 |@word mild:1 polynomial:18 norm:9 seems:1 nd:1 km:7 d2:2 covariance:1 decomposition:3 hsieh:1 incurs:1 thereby:1 versatile:1 series:2 contains:1 woodruff:9 com:2 si:2 written:1 truct:4 additive:14 numerical:2 confirming:1 dtq:3 designed:1 plot:2 hash:1 rrt:1 fewer:1 device:1 item:1 kyk:1 rudin:1 ksm:1 core:1 prov...
4,547
5,114
Trading Computation for Communication: Distributed Stochastic Dual Coordinate Ascent Tianbao Yang NEC Labs America, Cupertino, CA 95014 tyang@nec-labs.com Abstract We present and study a distributed optimization algorithm by employing a stochastic dual coordinate ascent method. Stochastic dual coordinate ascent method...
5114 |@word briefly:1 version:1 norm:7 open:1 hsieh:1 pressure:1 sgd:9 pick:2 mention:1 venkatasubramanian:1 initial:2 contains:1 past:1 bradley:1 com:2 comparing:4 luo:1 chu:2 xmk:1 kdd:7 update:10 bickson:1 v:4 selected:2 xk:8 beginning:2 core:4 provides:1 math:1 node:4 zhang:3 along:1 constructed:2 become:2 deterior...
4,548
5,115
Locally Adaptive Bayesian Multivariate Time Series Bruno Scarpa Department of Statistical Sciences University of Padua Via Cesare Battisti 241, 35121, Padua, Italy scarpa@stat.unipd.it Daniele Durante Department of Statistical Sciences University of Padua Via Cesare Battisti 241, 35121, Padua, Italy durante@stat.unip...
5115 |@word middle:1 version:1 inversion:2 polynomial:3 loading:2 seems:2 underline:1 confirms:2 simulation:13 covariance:29 dramatic:1 solid:4 accommodate:1 reduction:4 initial:2 series:19 united:1 denoting:1 longitudinal:1 reaction:1 com:2 worsening:1 hoboken:1 additive:1 enables:1 plot:8 treating:1 update:7 stationa...
4,549
5,116
A Latent Source Model for Nonparametric Time Series Classification George H. Chen MIT georgehc@mit.edu Stanislav Nikolov Twitter snikolov@twitter.com Devavrat Shah MIT devavrat@mit.edu Abstract For classifying time series, a nearest-neighbor approach is widely used in practice with performance often competitive with...
5116 |@word version:5 polynomial:1 nd:1 vldb:1 seek:1 prasad:2 accounting:2 decomposition:1 reduction:1 moment:2 initial:1 series:86 daniel:1 document:2 ours:1 batista:1 outperforms:2 existing:5 luigi:1 com:1 yet:2 guez:1 readily:1 luis:1 additive:1 christian:1 treating:1 plot:1 v:2 generative:1 half:3 fpr:6 contribute...
4,550
5,117
Multilinear Dynamical Systems for Tensor Time Series Mark Rogers Lei Li Stuart Russell EECS Department, University of California, Berkeley markrogersjr@berkeley.edu, {leili,russell}@cs.berkeley.edu Abstract Data in the sciences frequently occur as sequences of multidimensional arrays called tensors. How can hidden, ev...
5117 |@word mri:1 version:1 humidity:1 tensorial:1 calculus:1 covariance:9 decomposition:18 tr:7 recursively:1 reduction:2 series:25 contains:1 outperforms:1 existing:2 z2:1 yet:1 written:3 must:1 tenet:1 j1:13 compel:1 christian:1 motor:1 treating:1 update:2 aside:1 selected:1 ntrain:3 accordingly:1 isotropic:3 marine...
4,551
5,118
What do row and column marginals reveal about your dataset? Behzad Golshan Boston University behzad@cs.bu.edu John W. Byers Boston University byers@cs.bu.edu Evimaria Terzi Boston University evimaria@cs.bu.edu Abstract Numerous datasets ranging from group memberships within social networks to purchase histories on ...
5118 |@word version:1 polynomial:6 norm:1 open:1 cha:1 hyv:2 simulation:1 decomposition:1 p0:3 q1:1 pick:1 asks:1 recursively:1 reduction:1 liu:1 cobb:1 efficacy:2 selecting:1 interestingly:2 existing:4 recovered:1 comparing:2 yet:1 assigning:1 must:3 bie:1 john:1 informative:1 kdd:2 designed:1 plot:1 depict:1 ainen:1 ...
4,552
5,119
Error-Minimizing Estimates and Universal Entry-Wise Error Bounds for Low-Rank Matrix Completion Franz J. Kir?aly? Department of Statistical Science and Centre for Inverse Problems University College London f.kiraly@ucl.ac.uk Louis Theran? Institute of Mathematics Discrete Geometry Group Freie Universit?at Berlin thera...
5119 |@word multitask:1 version:1 briefly:1 polynomial:9 norm:7 manageable:1 km:15 closure:2 theran:4 covariance:2 ld:3 initial:1 contains:1 selecting:1 outperforms:2 yet:1 dx:2 exposing:1 determinantal:1 additive:3 e22:1 plot:1 acar:1 v:1 precaution:1 half:2 instantiate:1 parameterization:1 short:1 provides:1 math:2 o...
4,553
512
Structural Risk Minimization for Character Recognition I. Guyon, V. Vapnik, B. Boser, L. Bottou, and S. A. Solla AT&T Bell Laboratories Holmdel, NJ 07733, USA Abstract The method of Structural Risk Minimization refers to tuning the capacity of the classifier to the available amount of training data. This capacity is ...
512 |@word polynomial:2 advantageous:1 norm:1 nd:1 seems:1 duda:1 thereby:1 reduction:3 initial:1 contains:1 chervonenkis:1 diagonalized:1 wd:7 comparing:1 must:3 reminiscent:1 analytic:1 half:1 xk:2 te3t:3 provides:4 along:1 constructed:1 direct:1 become:2 c2:1 consists:1 introduce:2 expected:3 examine:1 brain:5 termi...
4,554
5,120
Synthesizing Robust Plans under Incomplete Domain Models Tuan A. Nguyen Subbarao Kambhampati Minh Do Arizona State University natuan@asu.edu Arizona State University rao@asu.edu NASA Ames Research Center minh.do@nasa.gov Abstract Most current planners assume complete domain models and focus on generating correct...
5120 |@word version:1 loading:4 d2:1 seek:1 pg:1 pick:4 mention:1 solid:1 delgado:1 reduction:1 initial:5 contains:3 ours:4 ala:1 prefix:1 past:1 existing:2 lave:1 current:2 yet:1 must:1 planet:1 distant:1 enables:1 intelligence:11 asu:2 amir:1 smith:1 completeness:1 provides:1 ames:1 five:1 c2:2 direct:1 become:1 init...
4,555
5,121
Which Space Partitioning Tree to Use for Search? A. G. Gray Georgia Tech. Atlanta, GA 30308 agray@cc.gatech.edu P. Ram Georgia Tech. / Skytree, Inc. Atlanta, GA 30308 p.ram@gatech.edu Abstract We consider the task of nearest-neighbor search with the class of binary-spacepartitioning trees, which includes kd-trees, p...
5121 |@word version:1 stronger:1 covariance:9 tr:1 harder:2 recursively:3 liu:1 contains:4 selecting:2 karger:1 mages:4 outperforms:1 existing:4 current:2 deteriorating:2 yet:1 partition:43 remove:2 hash:10 v:20 greedy:1 leaf:4 implying:1 intelligence:3 plane:2 xk:3 ruhl:1 farther:2 quantizer:2 quantized:1 node:21 trav...
4,556
5,122
Solving inverse problem of Markov chain with partial observations Tetsuro Morimura IBM Research - Tokyo tetsuro@jp.ibm.com Takayuki Osogami IBM Research - Tokyo osogami@jp.ibm.com Tsuyoshi Id?e IBM T.J. Watson Research Center tide@us.ibm.com Abstract The Markov chain is a convenient tool to represent the dynamics o...
5122 |@word version:2 logit:1 d2:2 seek:1 ld:10 initial:14 loc:6 score:2 past:1 existing:1 current:2 com:5 surprising:1 si:4 john:1 update:3 rd2:1 stationary:13 intelligence:5 device:1 accordingly:1 steepest:2 realizing:1 colored:1 readability:1 location:5 preference:3 org:1 simpler:1 zhang:1 mathematical:1 along:3 dif...
4,557
5,123
Robust Data-Driven Dynamic Programming Daniel Kuhn ?cole Polytechnique F?d?rale de Lausanne CH-1015 Lausanne, Switzerland daniel.kuhn@epfl.ch Grani A. Hanasusanto Imperial College London London SW7 2AZ, UK g.hanasusanto11@imperial.ac.uk Abstract In stochastic optimal control the distribution of the exogenous noise i...
5123 |@word mild:1 middle:1 polynomial:1 achievable:1 nd:1 d2:2 willing:1 simulation:1 seek:2 carolina:2 hu:1 incurs:1 thereby:1 tr:7 profit:6 harder:1 boundedness:1 ld:2 initial:1 minmax:1 series:2 exclusively:1 selecting:1 configuration:1 daniel:2 outperforms:2 existing:1 recovered:1 discretization:3 current:1 ka:1 y...
4,558
5,124
Online Variational Approximations to non-Exponential Family Change Point Models: With Application to Radar Tracking Ryan Turner Northrop Grumman Corp. ryan.turner@ngc.com Steven Bottone Northrop Grumman Corp. steven.bottone@ngc.com Clay Stanek Northrop Grumman Corp. clay.stanek@ngc.com Abstract The Bayesian online ...
5124 |@word aircraft:8 version:1 briefly:2 middle:2 km:4 p0:2 q1:1 incurs:1 mention:1 moment:2 liu:1 series:6 contains:2 pub:1 score:1 existing:2 current:2 com:3 nt:7 jupp:1 yet:2 must:6 john:1 partition:1 shape:2 analytic:1 grumman:3 treating:1 drop:2 update:14 n0:4 v:2 implying:1 generative:1 accordingly:1 inspection...
4,559
5,125
q-OCSVM: A q-Quantile Estimator for High-Dimensional Distributions Assaf Glazer Michael Lindenbaum Shaul Markovitch Department of Computer Science, Technion - Israel Institute of Technology {assafgr,mic,shaulm}@cs.technion.ac.il Abstract In this paper we introduce a novel method that can efficiently estimate a family...
5125 |@word repository:6 version:3 polynomial:1 proportion:4 smirnov:2 nd:1 decomposition:1 moment:1 liu:1 contains:1 series:2 document:8 ours:1 interestingly:1 outperforms:1 existing:2 recovered:2 scovel:1 surprising:2 noc:13 john:4 subsequent:1 informative:1 wellbehaved:1 analytic:1 bart:1 half:10 greedy:3 intelligen...
4,560
5,126
Unsupervised Structure Learning of Stochastic And-Or Grammars Kewei Tu Maria Pavlovskaia Song-Chun Zhu Center for Vision, Cognition, Learning and Art Departments of Statistics and Computer Science University of California, Los Angeles {tukw,mariapavl,sczhu}@ucla.edu Abstract Stochastic And-Or grammars compactly repres...
5126 |@word kgk:2 solan:1 decomposition:1 pick:2 accommodate:1 recursively:1 reduction:7 initial:7 configuration:18 contains:7 fragment:66 ours:6 o2:6 existing:7 outperforms:3 si:2 assigning:1 written:1 parsing:6 remove:1 update:3 alone:1 generative:1 selected:1 greedy:4 intelligence:4 accordingly:2 desktop:1 reciproca...
4,561
5,127
Rapid Distance-Based Outlier Detection via Sampling 1 Mahito Sugiyama1 Karsten M. Borgwardt1,2 Machine Learning and Computational Biology Research Group, MPIs T?ubingen, Germany 2 Zentrum f?ur Bioinformatik, Eberhard Karls Universit?at T?ubingen, Germany {mahito.sugiyama,karsten.borgwardt}@tuebingen.mpg.de Abstract ...
5127 |@word trial:3 repository:3 version:2 proportion:1 vldb:1 tried:1 dramatic:1 recursively:1 schwabacher:2 carry:1 moment:2 liu:2 zimek:3 score:19 lichman:1 renewed:1 interestingly:1 outperforms:7 spambase:3 current:1 surprising:1 si:5 numerical:1 partition:7 kdd:1 designed:1 half:1 ubuntu:1 pvldb:1 prize:1 core:2 r...
4,562
5,128
One-shot learning by inverting a compositional causal process Ruslan Salakhutdinov Dept. of Statistics and Computer Science University of Toronto rsalakhu@cs.toronto.edu Brenden M. Lake Dept. of Brain and Cognitive Sciences MIT brenden@mit.edu Joshua B. Tenenbaum Dept. of Brain and Cognitive Sciences MIT jbt@mit.edu...
5128 |@word trial:10 version:1 seems:2 open:3 instruction:2 tried:2 shot:15 initial:1 generatively:1 liu:1 score:4 zij:11 document:1 interestingly:1 current:1 com:1 si:19 yet:3 written:2 parsing:4 must:1 realistic:1 blur:2 shape:1 motor:11 designed:1 v:6 generative:3 selected:4 intelligence:6 item:2 plane:2 beginning:2...
4,563
5,129
Stochastic Majorization-Minimization Algorithms for Large-Scale Optimization Julien Mairal LEAR Project-Team - INRIA Grenoble julien.mairal@inria.fr Abstract Majorization-minimization algorithms consist of iteratively minimizing a majorizing surrogate of an objective function. Because of its simplicity and its wide a...
5129 |@word mild:1 stronger:2 norm:4 open:1 hu:1 hsieh:1 mention:1 tr:1 recursively:2 liblinear:16 initial:1 series:1 tuned:1 ours:1 past:2 existing:2 outperforms:1 current:4 comparing:1 tackling:1 written:2 fn:17 numerical:2 plot:1 update:10 n0:6 juditsky:1 stationary:7 website:1 beginning:1 core:2 iterates:4 provides...
4,564
513
Hierarchical Transformation of Space in the Visual System Alexandre Pouget Stephen A. Fisher Terrence J. Sejnowski Computational Neurobiology Laboratory The Salk Institute La Jolla, CA 92037 Abstract Neurons encoding simple visual features in area VI such as orientation, direction of motion and color are organized ...
513 |@word trial:1 schoen:1 trotter:1 simulation:1 initial:1 contains:2 john:2 shape:1 motor:1 zacks:1 designed:1 nonspatial:1 footing:1 contribute:1 location:15 successive:1 along:2 become:1 fixation:3 expected:1 simulator:1 provided:1 retinotopic:10 project:1 what:1 interpreted:2 monkey:1 recruit:1 developed:4 findin...
4,565
5,130
Robust Transfer Principal Component Analysis with Rank Constraints Yuhong Guo Department of Computer and Information Sciences Temple University, Philadelphia, PA 19122, USA yuhong@temple.edu Abstract Principal component analysis (PCA), a well-established technique for data analysis and processing, provides a convenien...
5130 |@word version:2 polynomial:1 norm:22 km:4 seek:4 decomposition:5 reduction:7 configuration:1 contains:2 efficacy:2 document:1 bc:7 outperforms:2 recovered:2 ka:4 nt:6 comparing:1 concatenate:1 additive:2 update:4 stationary:4 intelligence:1 discovering:2 provides:2 zhang:1 mathematical:1 rnt:4 constructed:3 ik:2 ...
4,566
5,131
Online Robust PCA via Stochastic Optimization Huan Xu ME Department National University of Singapore mpexuh@nus.edu.sg Jiashi Feng ECE Department National University of Singapore jiashi@nus.edu.sg Shuicheng Yan ECE Department National University of Singapore eleyans@nus.edu.sg Abstract Robust PCA methods are typica...
5131 |@word mild:1 version:1 briefly:1 norm:16 shuicheng:1 simulation:4 covariance:1 decomposition:4 accounting:1 ronchetti:1 tr:2 klk:1 reduction:1 initial:1 mpexuh:1 contains:1 ours:1 past:1 recovered:2 current:2 comparing:1 luo:1 pcp:20 john:2 fn:2 realistic:2 additive:1 numerical:1 plot:5 update:5 stationary:3 plan...
4,567
5,132
The Fast Convergence of Incremental PCA Akshay Balsubramani UC San Diego abalsubr@cs.ucsd.edu Sanjoy Dasgupta UC San Diego dasgupta@cs.ucsd.edu Yoav Freund UC San Diego yfreund@cs.ucsd.edu Abstract We consider a situation in which we see samples Xn ? Rd drawn i.i.d. from some distribution with mean zero and unknown...
5132 |@word mild:2 version:1 seems:2 norm:1 stronger:1 open:1 d2:3 crucially:1 covariance:8 pick:3 sgd:2 vno:12 moment:5 necessity:1 reduction:1 series:1 exclusively:1 initial:6 written:1 fn:14 subsequent:1 additive:1 drop:2 update:10 n0:1 generative:2 prohibitive:1 intelligence:2 warmuth:2 oldest:1 characterization:1 ...
4,568
5,133
Probabilistic Principal Geodesic Analysis P. Thomas Fletcher School of Computing University of Utah Salt Lake City, UT fletcher@sci.utah.edu Miaomiao Zhang School of Computing University of Utah Salt Lake City, UT miaomiao@sci.utah.edu Abstract Principal geodesic analysis (PGA) is a generalization of principal compo...
5133 |@word determinant:2 mri:3 briefly:2 middle:1 norm:1 open:1 covariance:1 reduction:1 initial:6 configuration:2 contains:1 series:1 zij:7 rkhs:1 current:2 dx:1 must:2 written:1 john:1 numerical:2 shape:23 remove:1 atlas:1 update:2 generative:1 intelligence:2 accordingly:1 plane:1 isotropic:2 hamiltonian:5 smith:1 h...
4,569
5,134
Fast Algorithms for Gaussian Noise Invariant Independent Component Analysis Luis Rademacher James Voss Ohio State University Ohio State University Computer Science and Engineering, Computer Science and Engineering, 2015 Neil Avenue, Dreese Labs 495. 2015 Neil Avenue, Dreese Labs 586. Columbus, OH 43210 Columbus, OH 43...
5134 |@word briefly:1 version:5 polynomial:1 nd:1 d2:2 hu:8 simulation:1 hyv:4 covariance:6 decomposition:2 decorrelate:1 mlk:1 atrix:2 moment:7 series:1 interestingly:1 outperforms:1 existing:1 comparing:1 surprising:1 si:8 written:1 luis:1 additive:6 numerical:1 subsequent:1 wx:1 update:5 generative:1 intelligence:1 ...
4,570
5,135
Online PCA for Contaminated Data Jiashi Feng ECE Department National University of Singapore jiashi@nus.edu.sg Huan Xu ME Department National University of Singapore mpexuh@nus.edu.sg Shie Mannor EE Department Technion shie@ee.technion.ac.il Shuicheng Yan ECE Department National University of Singapore eleyans@nus.e...
5135 |@word mild:3 determinant:1 briefly:1 norm:2 c0:2 shuicheng:1 simulation:5 covariance:7 decomposition:2 ronchetti:1 minus:1 initial:19 mpexuh:1 contains:1 ours:2 current:5 wd:2 scatter:1 must:1 john:3 distant:1 partition:3 numerical:1 remove:1 drop:1 update:8 intelligence:1 selected:1 warmuth:1 accordingly:3 plane...
4,571
5,136
Fantope Projection and Selection: A near-optimal convex relaxation of sparse PCA Vincent Q. Vu The Ohio State University vqv@stat.osu.edu Juhee Cho University of Wisconsin, Madison chojuhee@stat.wisc.edu Jing Lei Carnegie Mellon University leij09@gmail.com Karl Rohe University of Wisconsin, Madison karlrohe@stat.wis...
5136 |@word version:3 polynomial:2 seems:1 norm:11 open:1 simulation:5 covariance:17 decomposition:2 tr:4 reduction:2 initial:1 plentiful:1 uncovered:1 dspca:6 liu:3 ours:2 nonparanormal:3 past:1 existing:2 elliptical:2 com:1 comparing:2 gmail:1 scatter:1 yet:1 subsequent:1 numerical:1 weyl:2 statis:6 update:3 mackey:1...
4,572
5,137
One-shot learning and big data with n = 2 Dean P. Foster University of Pennsylvania Philadelphia, PA dean@foster.net Lee H. Dicker Rutgers University Piscataway, NJ ldicker@stat.rutgers.edu Abstract We model a ?one-shot learning? situation, where very few observations y1 , ..., yn ? R are available. Associated with ...
5137 |@word version:1 nd:1 open:1 d2:1 simulation:5 covariance:2 shot:27 moment:2 necessity:1 contains:1 score:3 series:3 neeman:1 bc:36 o2:1 existing:1 outperforms:3 contextual:3 comparing:2 subsequent:1 numerical:1 informative:2 enables:1 pursued:2 intelligence:1 warmuth:1 inspection:1 beginning:1 vanishing:1 smith:1...
4,573
5,138
The Randomized Dependence Coefficient David Lopez-Paz, Philipp Hennig, Bernhard Sch?olkopf Max Planck Institute for Intelligent Systems Spemannstra?e 38, T?ubingen, Germany {dlopez,phennig,bs}@tue.mpg.de Abstract We introduce the Randomized Dependence Coefficient (RDC), a measure of nonlinear dependence between rando...
5138 |@word version:2 reshef:4 norm:1 nd:2 crucially:1 decomposition:1 kent:1 covariance:1 nystr:1 series:2 score:2 selecting:3 lightweight:1 favouring:1 analysed:1 si:2 must:1 additive:3 numerical:1 partition:2 designed:1 plot:1 statis:1 n0:1 greedy:2 selected:1 sarcos:2 detecting:1 philipp:1 org:2 zhang:1 five:2 dn:1...
4,574
5,139
Sparse Additive Text Models with Low Rank Background Lei Shi Baidu.com, Inc. P.R. China shilei06@baidu.om Abstract The sparse additive model for text modeling involves the sum-of-exp computing, whose cost is consuming for large scales. Moreover, the assumption of equal background across all classes/topics may be too ...
5139 |@word version:1 middle:1 advantageous:1 norm:12 proportion:5 seek:1 linearized:1 decomposition:2 eng:1 thres:1 pick:1 inefficiency:1 loc:1 denoting:1 document:19 interestingly:3 outperforms:1 existing:3 current:1 com:1 yet:2 assigning:3 tackling:1 additive:16 plot:2 update:8 generative:11 selected:1 discovering:4...
4,575
514
Hierarchies of adaptive experts Robert A. Jacobs Michael I. Jordan Department of Brain and Cognitive Sciences Massachusetts Institute of Technology Cambridge, MA 02139 Abstract In this paper we present a neural network architecture that discovers a recursive decomposition of its input space. Based on a generalization...
514 |@word autoassociator:1 seems:1 simulation:2 lobe:1 jacob:8 decomposition:3 covariance:1 recursively:2 trinary:1 comparing:1 nowlan:4 activation:3 si:1 yet:1 must:1 readily:1 realize:1 belmont:1 additive:1 partition:4 treating:1 selected:1 leaf:6 item:1 flare:1 ith:8 coarse:2 parameterizations:1 ames:1 oak:3 mathem...
4,576
5,140
Documents as multiple overlapping windows into a grid of counts Alessandro Perina1 Nebojsa Jojic1 1 Manuele Bicego2 Andrzej Turski1 Microsoft Corporation, Redmond, WA 2 University of Verona, Italy Abstract In text analysis documents are often represented as disorganized bags of words; models of such count features...
5140 |@word middle:2 kondor:1 proportion:3 verona:1 seal:2 almond:2 bun:1 instruction:1 gradual:1 chili:2 pick:9 reduction:2 wrapper:1 rind:1 moment:1 score:2 slotted:1 yxx:1 salzmann:1 document:24 shrimp:1 interestingly:1 outperforms:1 wd:3 cooker:1 surprising:1 com:1 must:2 grain:1 refines:1 distant:1 visible:4 shape...
4,577
5,141
On Algorithms for Sparse Multi-factor NMF Siwei Lyu Xin Wang Computer Science Department University at Albany, SUNY Albany, NY 12222 {slyu,xwang26}@albany.edu Abstract Nonnegative matrix factorization (NMF) is a popular data analysis method, the objective of which is to approximate a matrix with all nonnegative com...
5141 |@word illustrating:1 norm:9 seek:2 decomposition:6 brightness:1 solid:1 edric:1 reduction:1 electronics:1 configuration:1 contains:2 initial:4 daniel:1 document:3 interestingly:1 outperforms:1 freitas:1 comparing:1 subsequent:1 numerical:3 remove:1 update:13 n0:20 v:1 greedy:2 fewer:1 prohibitive:1 xk:27 ugander:...
4,578
5,142
Learning Adaptive Value of Information for Structured Prediction Ben Taskar University of Washington Seattle, WA taskar@cs.washington.edu David Weiss University of Pennsylvania Philadelphia, PA djweiss@cis.upenn.edu Abstract Discriminative methods for learning structured models have enabled wide-spread use of very ri...
5142 |@word version:1 eliminating:1 pick:1 recursively:1 initial:1 liu:1 series:1 score:4 tuned:1 ours:2 past:1 current:8 nt:1 si:3 yet:1 assigning:1 must:4 parsing:4 concatenate:2 informative:2 hofmann:1 sponsored:1 update:2 resampling:1 generative:1 prohibitive:1 cue:1 selected:1 greedy:9 intelligence:1 mccallum:1 sh...
4,579
5,143
Symbolic Opportunistic Policy Iteration for Factored-Action MDPs Aswin Raghavana Roni Khardonb Alan Ferna Prasad Tadepallia a School of EECS, Oregon State University, Corvallis, OR, USA {nadamuna,afern,tadepall}@eecs.orst.edu b Department of Computer Science, Tufts University, Medford, MA, USA roni@cs.tufts.edu Abstra...
5143 |@word version:4 compression:5 coarseness:1 tadepalli:1 heuristically:1 hu:1 prasad:2 delgado:1 initial:1 interestingly:1 existing:1 current:4 written:1 must:1 gaona:1 john:1 ronald:2 treating:1 plot:1 update:2 v:2 greedy:4 leaf:8 prohibitive:1 intelligence:3 offpolicy:1 core:2 provides:1 node:7 karina:1 successiv...
4,580
5,144
Point Based Value Iteration with Optimal Belief Compression for Dec-POMDPs Charles L. Isbell College of Computing Georgia Institute of Technology Atlanta, GA 30332 isbell@cc.gatech.edu Liam MacDermed College of Computing Georgia Institute of Technology Atlanta, GA 30332 liam@cc.gatech.edu Abstract We present four ma...
5144 |@word version:1 compression:17 open:3 contraction:1 concise:1 recursively:2 initial:4 series:1 selecting:1 outperforms:3 existing:11 current:10 skipping:1 must:10 partition:1 designed:1 n0:1 greedy:1 selected:1 fewer:1 intelligence:9 core:1 mental:1 revisited:1 successive:1 hyperplanes:3 traverse:1 five:3 along:7...
4,581
5,145
Convergence of Monte Carlo Tree Search in Simultaneous Move Games Viliam Lis?y1 Vojt?ech Kova?r??k1 Marc Lanctot2 Branislav Bo?sansk?y1 2 1 Department of Knowledge Engineering Maastricht University, The Netherlands marc.lanctot @maastrichtuniversity.nl Agent Technology Center Dept. of Computer Science and Engine...
5145 |@word version:2 achievable:1 polynomial:1 bf:5 d2:1 simulation:9 maes:1 initial:2 contains:1 selecting:4 denoting:1 current:7 comparing:1 michal:2 assigning:1 numerical:1 happen:1 j1:1 update:9 intelligence:5 selected:8 advancement:1 leaf:10 vmin:3 serialized:1 short:1 infrastructure:2 node:19 org:1 teytaud:1 mat...
4,582
5,146
Estimation Bias in Multi-Armed Bandit Algorithms for Search Advertising Tao Qin Microsoft Research Asia taoqin@microsoft.com Min Xu Machine Learning Department Carnegie Mellon University minx@cs.cmu.edu Tie-Yan Liu Microsoft Research Asia tie-yan.liu@microsoft.com Abstract In search advertising, the search engine n...
5146 |@word trial:1 exploitation:1 briefly:1 version:1 seems:1 stronger:1 c0:2 unif:7 willing:1 simulation:6 bn:4 reduction:1 liu:2 score:3 selecting:1 omniscient:1 existing:1 current:1 com:2 contextual:1 must:2 realistic:1 numerical:1 partition:1 alam:1 drop:1 sponsored:5 update:1 half:1 fewer:2 record:2 vorobeychik:1...
4,583
5,147
Optimization, Learning, and Games with Predictable Sequences Alexander Rakhlin University of Pennsylvania Karthik Sridharan University of Pennsylvania Abstract We provide several applications of Optimistic Mirror Descent, an online learning algorithm based on the idea of predictable sequences. First, we recover the M...
5147 |@word version:4 norm:8 nd:1 open:2 unif:4 gradual:1 decomposition:1 asks:1 dramatic:1 reduction:1 selecting:1 interestingly:1 past:1 nt:10 yet:2 benign:1 remove:1 designed:1 hypothesize:1 update:8 fund:1 guess:1 beginning:2 chiang:1 provides:1 equi:1 contribute:1 simpler:1 kelner:1 mathematical:1 h4:2 direct:1 sy...
4,584
5,148
Minimax Optimal Algorithms for Unconstrained Linear Optimization H. Brendan McMahan Google Reasearch Seattle, WA mcmahan@google.com Jacob Abernethy? Computer Science and Engineering University of Michigan jabernet@umich.edu Abstract We design and analyze minimax-optimal algorithms for online linear optimization games...
5148 |@word version:1 polynomial:1 norm:1 stronger:1 replicate:2 dekel:2 additively:1 jacob:6 citeseer:1 mention:1 minus:1 initial:2 contains:2 selecting:1 egt:1 interestingly:1 existing:1 com:1 surprising:1 yet:1 must:5 written:1 update:1 warmuth:4 manfred:3 characterization:8 provides:1 earnings:1 gx:4 mathematical:1...
4,585
5,149
Online Learning with Costly Features and Labels Navid Zolghadr Department of Computing Science University of Alberta zolghadr@ualberta.ca G?abor Bart?ok Department of Computer Science ETH Z?urich bartok@inf.ethz.ch Russell Greiner Andr?as Gy?orgy Csaba Szepesv?ari Department of Computing Science, University of Albert...
5149 |@word mild:1 innovates:1 version:6 achievable:1 norm:2 seems:2 dekel:2 open:3 contains:1 selecting:3 denoting:1 existing:1 current:2 discretization:6 nt:8 si:2 yet:1 must:3 readily:2 realistic:1 drop:2 update:1 bart:2 intelligence:1 selected:4 device:1 beginning:1 ith:2 prespecified:2 record:1 d2d:1 provides:3 eq...
4,586
515
Linear Operator for Object Recognition Ronen Bssri Shimon Ullman? M.I.T. Artificial Intelligence Laboratory and Department of Brain and Cognitive Science 545 Technology Square Cambridge, MA 02139 Abstract Visual object recognition involves the identification of images of 3-D objects seen from arbitrary viewpoints. We...
515 |@word version:1 middle:1 solid:1 contains:2 existing:2 xand:1 assigning:1 must:1 visible:1 shape:2 designed:1 intelligence:5 core:1 math:1 location:3 simpler:1 constructed:1 supply:1 edelman:2 consists:1 prove:1 fitting:1 recognizable:1 introduce:2 commenting:1 lehtio:2 brain:1 compensating:1 actual:2 considering:...
4,587
5,150
The Pareto Regret Frontier Wouter M. Koolen Queensland University of Technology wouter.koolen@qut.edu.au Abstract Performance guarantees for online learning algorithms typically take the form of regret bounds, which express that the cumulative loss overhead compared to the best expert in hindsight is small. In the co...
5150 |@word trial:1 briefly:1 manageable:1 achievable:8 seems:1 compression:1 version:2 open:1 gradual:1 queensland:1 jacob:3 incurs:1 mention:1 harder:1 substitution:1 series:1 daniel:1 tuned:1 erven:1 comparing:1 analysed:2 must:3 john:1 additive:1 progressively:2 update:2 v:2 pursued:1 warmuth:8 core:1 manfred:4 chi...
4,588
5,151
Online Learning with Switching Costs and Other Adaptive Adversaries Nicol`o Cesa-Bianchi Universit`a degli Studi di Milano Italy Ofer Dekel Microsoft Research USA Ohad Shamir Microsoft Research and the Weizmann Institute Abstract We study the power of different types of adaptive (nonoblivious) adversaries in the set...
5151 |@word exploitation:4 version:5 stronger:2 dekel:2 open:1 seek:1 attainable:3 invoking:1 eld:1 incurs:1 arti:1 harder:1 reduction:8 ours:1 interestingly:1 past:5 existing:2 current:4 od:1 surprising:1 written:1 must:5 informative:1 cant:1 designed:2 interpretable:1 implying:2 intelligence:1 warmuth:1 accordingly:1...
4,589
5,152
High-Dimensional Gaussian Process Bandits Josip Djolonga ETH Z?urich josipd@ethz.ch Andreas Krause ETH Z?urich krausea@ethz.ch Volkan Cevher EPFL volkan.cevher@epfl.ch Abstract Many applications in machine learning require optimizing unknown functions defined over a high-dimensional space from noisy samples that ar...
5152 |@word determinant:1 exploitation:9 middle:1 polynomial:1 norm:12 faculty:1 nd:1 suitably:2 km:1 d2:1 decomposition:1 pick:5 dramatic:1 incurs:1 asks:1 thereby:1 tr:1 harder:1 carry:1 reduction:1 score:5 rkhs:12 outperforms:3 existing:1 freitas:2 ka:3 si:17 bd:3 must:2 lorentz:1 numerical:2 confirming:1 shape:1 re...
4,590
5,153
On Poisson Graphical Models Eunho Yang Department of Computer Science University of Texas at Austin eunho@cs.utexas.edu Pradeep Ravikumar Department of Computer Science University of Texas at Austin pradeepr@cs.utexas.edu Genevera I. Allen Department of Statistics and Electrical & Computer Engineering Rice Universit...
5153 |@word briefly:1 version:1 middle:1 seems:2 simulation:1 covariance:1 accommodate:1 liu:8 series:3 denoting:1 document:1 interestingly:1 nonparanormal:2 genetic:1 existing:2 comparing:1 incidence:3 yet:1 written:1 reminiscent:1 must:1 numerical:1 partition:8 shape:1 atlas:2 depict:1 accordingly:3 sys:1 caveat:1 no...
4,591
5,154
Conditional Random Fields via Univariate Exponential Families Eunho Yang Department of Computer Science University of Texas at Austin eunho@cs.utexas.edu Pradeep Ravikumar Department of Computer Science University of Texas at Austin pradeepr@cs.utexas.edu Genevera I. Allen Department of Statistics and Electrical & Co...
5154 |@word trial:2 cu:2 version:1 c0:1 tensorial:1 seek:2 simulation:4 covariance:9 tr:6 t2n:3 liu:7 series:1 egfr:3 denoting:1 genetic:2 outperforms:1 existing:2 bradley:2 contextual:1 bsj:2 dx:3 must:1 attracted:1 written:1 partition:5 plot:2 atlas:2 v:1 alone:1 generative:3 greedy:1 selected:2 accordingly:1 paramet...
4,592
5,155
Scalable kernels for graphs with continuous attributes Aasa Feragen, Niklas Kasenburg Machine Learning and Computational Biology Group Max Planck Institutes T?ubingen and DIKU, University of Copenhagen {aasa,niklas.kasenburg}@diku.dk Jens Petersen1 , Marleen de Bruijne1,2 1 DIKU, University of Copenhagen 2 Erasmus Med...
5155 |@word trial:1 version:1 middle:1 kondor:1 flach:1 open:2 dirksen:4 recursively:3 contains:2 perret:1 outperforms:1 ka:1 comparing:4 must:2 cruz:1 shape:1 gv:15 plot:1 update:3 hash:2 leaf:2 selected:1 ith:5 prize:1 short:1 node:90 height:1 mehlhorn:2 along:7 schweitzer:1 become:1 transducer:1 prove:2 consists:1 d...
4,593
5,156
Near-optimal Anomaly Detection in Graphs using Lov?asz Extended Scan Statistic Akshay Krishnamurthy Computer Science Department Carnegie Mellon University Pittsburgh, PA 15213 akshaykr@cs.cmu.edu James Sharpnack Machine Learning Department Carnegie Mellon University Pittsburgh, PA 15213 jsharpna@gmail.com Aarti Sing...
5156 |@word mild:1 briefly:1 kondor:2 polynomial:1 norm:2 physik:1 calculus:1 bn:3 decomposition:3 jacob:1 citeseer:1 commute:1 dramatic:1 venkatasubramanian:1 selecting:1 daniel:1 ours:1 document:1 kahl:1 aberrant:1 com:1 incidence:1 activation:5 gmail:1 must:3 written:1 ust:5 stemming:1 john:1 ronald:1 limp:1 j1:3 an...
4,594
5,157
Analyzing the Harmonic Structure in Graph-Based Learning Xiao-Ming Wu1 , Zhenguo Li3 , and Shih-Fu Chang1,2 1 Department of Electrical Engineering, Columbia University 2 Department of Computer Science, Columbia University 3 Huawei Noah?s Ark Lab, Hong Kong {xmwu, sfchang}@ee.columbia.edu, li.zhenguo@huawei.com Abstra...
5157 |@word kong:1 trial:1 stronger:1 proportion:1 seems:3 open:2 d2:1 confirms:1 simulation:1 commute:6 ld:1 carry:1 interestingly:1 existing:1 com:1 comparing:6 si:54 lang:1 attracted:1 must:1 chicago:1 informative:3 drop:7 a1k:4 implying:1 provides:3 node:1 five:3 mathematical:4 dn:2 become:1 focs:1 consists:1 coifm...
4,595
5,158
Learning Gaussian Graphical Models with Observed or Latent FVSs Alan S. Willsky Department of EECS Massachusetts Institute of Technology willsky@mit.edu Ying Liu Department of EECS Massachusetts Institute of Technology liu_ying@mit.edu Abstract Gaussian Graphical Models (GGMs) or Gauss Markov random fields are widely...
5158 |@word determinant:1 version:3 briefly:1 manageable:1 polynomial:6 open:1 covariance:24 decomposition:1 concise:1 reduction:2 initial:4 liu:22 contains:1 series:1 selecting:1 karger:1 current:1 recovered:1 si:2 dx:1 chicago:1 partition:4 jfk:1 plot:1 alone:1 greedy:9 selected:6 intelligence:1 parametrization:1 fa9...
4,596
5,159
Global MAP-Optimality by Shrinking the Combinatorial Search Area with Convex Relaxation Bogdan Savchynskyy1 J?org Kappes2 Paul Swoboda2 Christoph Schn?orr1,2 1 Heidelberg Collaboratory for Image Processing, Heidelberg University, Germany bogdan.savchynskyy@iwr.uni-heidelberg.de 2 Image and Pattern Analysis Group, Heide...
5159 |@word version:1 manageable:1 stronger:1 norm:2 flach:1 tried:1 grk:1 decomposition:6 inpainting:1 carry:1 n8:3 initial:11 contains:4 efficacy:1 series:1 past:1 existing:1 com:1 givry:1 dechter:1 numerical:1 remove:1 drop:1 greedy:1 selected:1 intelligence:1 plane:9 smith:1 recompute:1 math:1 node:33 provides:3 re...
4,597
516
Neural Network Routing for Random Multistage Interconnection Networks Mark W. Goudreau Princeton University and NEe Research Institute, Inc. 4 Independence Way Princeton, NJ 08540 c. Lee Giles NEC Research Institute, Inc. 4 Independence Way Princeton, NJ 08540 Abstract A routing scheme that uses a neural network has...
516 |@word simulation:2 accommodate:1 ld:2 liu:2 mag:1 suppressing:1 current:4 router:37 must:2 oml:1 subsequent:1 designed:1 greedy:11 selected:1 liapunov:1 beginning:2 sys:1 provides:1 successive:1 rc:1 constructed:1 viable:1 manner:2 expected:2 indeed:1 behavior:1 globally:1 encouraging:1 actual:1 considering:1 incr...
4,598
5,160
First-Order Decomposition Trees Nima Taghipour Jesse Davis Hendrik Blockeel Department of Computer Science, KU Leuven Celestijnenlaan 200A, B-3001 Heverlee, Belgium Abstract Lifting attempts to speedup probabilistic inference by exploiting symmetries in the model. Exact lifted inference methods, like their propositio...
5160 |@word version:2 polynomial:2 dtrees:28 vi1:1 nd:3 adnan:2 decomposition:34 accounting:1 innermost:1 minus:1 accommodate:1 recursively:4 substitution:4 configuration:3 contains:4 interestingly:1 existing:3 contextual:1 nt:11 si:1 assigning:1 dx:6 written:4 dechter:1 partition:2 plm:7 fund:1 alone:1 intelligence:12...
4,599
5,161
Binary to Bushy: Bayesian Hierarchical Clustering with the Beta Coalescent Yuening Hu1 , Jordan Boyd-Graber2 , Hal Daum`e III3 , Z. Irene Ying4 1, 3: Computer Science, 2: iSchool and UMIACS, 4: Agricultural Research Service 1, 2, 3: University of Maryland, 4: Department of Agriculture ynhu@cs.umd.edu, {jbg,hal}@umiacs....
5161 |@word briefly:1 c0:2 covariance:2 p0:5 recursively:1 initial:6 contains:1 score:15 document:6 dpmms:4 outperforms:2 existing:3 freitas:1 recovered:2 com:1 comparing:1 si:4 must:2 parsing:2 john:1 subsequent:2 partition:19 realistic:1 christian:2 remove:1 atlas:1 interpretable:1 update:4 resampling:1 half:1 discov...