Unnamed: 0
int64
0
7.24k
id
int64
1
7.28k
raw_text
stringlengths
9
124k
vw_text
stringlengths
12
15k
6,300
67
602 GENERALIZATION OF BACKPROPAGATION TO RECURRENT AND HIGHER ORDER NEURAL NETWORKS Fernando J. Pineda Applied Physics Laboratory, Johns Hopkins University Johns Hopkins Rd., Laurel MD 20707 Abstract A general method for deriving backpropagation algorithms for networks with recurrent and higher order networks is intr...
67 |@word version:1 inversion:2 polynomial:1 tedious:1 r:1 linearized:1 simulation:3 fmite:1 fonn:5 jacqueline:1 moment:1 initial:8 contains:1 lapedes:4 activation:2 dx:3 must:5 luis:1 john:3 numerical:2 visible:5 update:9 stationary:1 tenn:1 alone:1 yr:3 liapunov:2 xk:2 ji2:1 draft:1 location:1 ofbackpropagation:1 mat...
6,301
670
A Model of Feedback to the Lateral Geniculate Nucleus Carlos D. Brody Computation and Neural Systems Program California Institute of Technology Pasadena, CA 91125 Abstract Simplified models of the lateral geniculate nucles (LGN) and striate cortex illustrate the possibility that feedback to the LG N may be used for ro...
670 |@word illustrating:1 briefly:1 wiesel:1 cha:1 grey:1 simulation:2 fortuitous:1 phy:1 contains:1 tuned:7 rightmost:2 activation:2 atop:1 must:2 physiol:1 alone:1 isotropic:1 short:3 farther:2 compo:1 correlat:1 detecting:7 provides:1 location:1 preference:1 along:1 direct:2 pathway:5 combine:1 brain:4 unfolded:1 ac...
6,302
6,700
SchNet: A continuous-filter convolutional neural network for modeling quantum interactions ? K. T. Sch?tt1?, P.-J. Kindermans1 , H. E. Sauceda2 , S. Chmiela1 A. Tkatchenko3 , K.-R. M?ller1,4,5? 1 Machine Learning Group, Technische Universit?t Berlin, Germany 2 Theory Department, Fritz-Haber-Institut der Max-Planck-Ge...
6700 |@word middle:2 briefly:1 kondor:2 nd:1 simulation:3 sgd:1 initial:2 configuration:1 contains:2 series:3 outperforms:2 o2:3 current:1 activation:2 fn:1 shape:1 enables:1 designed:4 update:3 bart:2 alone:1 generative:1 fewer:1 leaf:1 beginning:1 hamiltonian:1 vanishing:1 short:2 provides:1 location:2 org:1 zhang:1 ...
6,303
6,701
Active Bias: Training More Accurate Neural Networks by Emphasizing High Variance Samples Haw-Shiuan Chang, Erik Learned-Miller, Andrew McCallum University of Massachusetts, Amherst 140 Governors Dr., Amherst, MA 01003 {hschang,elm,mccallum}@cs.umass.edu Abstract Self-paced learning and hard example mining re-weight t...
6701 |@word trial:5 exploitation:1 version:3 cnn:16 nd:2 tried:2 bn:1 sgd:104 harder:5 reduction:6 initial:2 liu:2 lightweight:3 uma:1 selecting:6 score:1 seriously:1 document:1 fa8750:1 existing:2 freitas:1 current:4 wd:9 nt:3 com:6 informative:2 shape:1 drop:2 sponsored:1 moczulski:1 selected:3 mccallum:5 beginning:3...
6,304
6,702
Differentiable Learning of Submodular Models Andreas Krause Department of Computer Science ETH Zurich krausea@ethz.ch Josip Djolonga Department of Computer Science ETH Zurich josipd@inf.ethz.ch Abstract Can we incorporate discrete optimization algorithms within modern machine learning models? For example, is it poss...
6702 |@word kohli:1 cnn:6 polynomial:5 seems:4 norm:26 open:3 checkable:1 seek:1 rgb:2 pick:1 sgd:1 kijima:1 reduction:1 configuration:3 series:1 score:4 com:1 intriguing:1 must:1 subsequent:1 partition:5 hofmann:1 remove:1 dolhansky:2 fewer:2 selected:1 intelligence:3 xk:1 parametrization:1 certificate:1 characterizat...
6,305
6,703
Inductive Representation Learning on Large Graphs William L. Hamilton? wleif@stanford.edu Rex Ying? rexying@stanford.edu Jure Leskovec jure@cs.stanford.edu Department of Computer Science Stanford University Stanford, CA, 94305 Abstract Low-dimensional embeddings of nodes in large graphs have proved extremely usefu...
6703 |@word trial:2 repository:1 version:4 proportion:1 tamayo:1 propagate:1 kutzkov:1 sgd:2 reduction:1 contains:6 score:5 daniel:1 document:1 interestingly:1 dubourg:1 outperforms:7 existing:2 current:3 comparing:2 activation:2 si:2 must:2 devin:1 concatenate:1 kdd:4 designed:4 hash:2 alone:1 v:1 advancement:2 concat...
6,306
6,704
Subset Selection and Summarization in Sequential Data Ehsan Elhamifar Computer and Information Science College Northeastern University Boston, MA 02115 eelhami@ccs.neu.edu M. Clara De Paolis Kaluza Computer and Information Science College Northeastern University Boston, MA 02115 clara@ccs.neu.edu Abstract Subset sele...
6704 |@word polynomial:1 compression:5 duda:1 open:1 instruction:1 seitz:1 decomposition:1 elisseeff:1 pick:1 reduction:1 initial:1 series:4 score:20 selecting:13 loeliger:1 denoting:3 document:6 outperforms:2 existing:3 disaggregation:1 clara:2 assigning:1 chu:1 written:1 determinantal:5 partition:1 informative:1 remo...
6,307
6,705
Question Asking as Program Generation Anselm Rothe1 anselm@nyu.edu 1 Brenden M. Lake1,2 brenden@nyu.edu Todd M. Gureckis1 todd.gureckis@nyu.edu 2 Department of Psychology Center for Data Science New York University Abstract A hallmark of human intelligence is the ability to ask rich, creative, and revealing questi...
6705 |@word polynomial:1 laurence:1 open:4 instruction:1 calculus:1 concise:2 tmg:1 recursively:1 reduction:1 configuration:5 contains:2 score:1 bootstrapped:1 prefix:1 past:1 current:5 com:1 si:7 yet:1 must:2 john:4 numerical:1 partition:2 informative:4 shape:1 wanted:1 remove:1 designed:3 interpretable:1 update:1 dro...
6,308
6,706
Revisiting Perceptron: Ef?cient and Label-Optimal Learning of Halfspaces Songbai Yan UC San Diego La Jolla, CA yansongbai@ucsd.edu Chicheng Zhang? Microsoft Research New York, NY chicheng.zhang@microsoft.com Abstract It has been a long-standing problem to ef?ciently learn a halfspace using as few labels as possible i...
6706 |@word polynomial:10 nd:4 dekel:1 open:4 d2:14 hu:7 km:3 prasad:1 arti:2 harder:1 initial:10 liu:2 series:1 chervonenkis:1 daniel:2 ours:2 existing:1 err:13 current:1 com:1 kmk:1 beygelzimer:3 dx:9 must:1 john:5 informative:1 hongyang:3 cheap:1 drop:1 atlas:1 update:3 intelligence:2 selected:1 half:1 isotropic:1 b...
6,309
6,707
Gradient Descent Can Take Exponential Time to Escape Saddle Points Chi Jin University of California, Berkeley chijin@berkeley.edu Simon S. Du Carnegie Mellon University ssdu@cs.cmu.edu Jason D. Lee University of Southern California jasonlee@marshall.usc.edu Michael I. Jordan University of California, Berkeley jordan@...
6707 |@word version:3 polynomial:20 stronger:1 norm:2 nd:8 open:2 unif:1 willing:1 simulation:1 decomposition:2 covariance:2 arti:2 thereby:2 necessity:2 liu:1 daniel:1 xinyang:1 fa8750:1 past:1 existing:1 current:2 luo:2 naman:1 written:2 must:2 readily:1 john:5 analytic:1 plot:2 update:2 stationary:20 intelligence:2 ...
6,310
6,708
Union of Intersections (UoI) for Interpretable Data Driven Discovery and Prediction Kristofer E. Bouchard? Alejandro F. Bujan? Shashanka Ubaru? Prabhat? Edward F. Chang?? Michael W. Mahoney? Farbod Roosta-Khorasani? Antoine M. Snijdersk Jian-Hua Maok Sharmodeep Bhattacharyya?? Abstract The increasing size an...
6708 |@word trial:1 version:1 compression:7 norm:2 grey:1 seek:1 decomposition:6 thereby:1 solid:1 initial:1 series:1 genetic:3 bootstrapped:1 interestingly:1 bhattacharyya:2 existing:2 ka:1 com:1 gmail:1 scatter:1 numerical:3 plasticity:1 motor:1 plot:4 interpretable:5 resampling:5 v:3 generative:1 selected:4 fewer:2 ...
6,311
6,709
One-Shot Imitation Learning Yan Duan?? , Marcin Andrychowicz? , Bradly Stadie?? , Jonathan Ho?? , Jonas Schneider? , Ilya Sutskever? , Pieter Abbeel?? , Wojciech Zaremba? ? Berkeley AI Research Lab, ? OpenAI ? Work done while at OpenAI {rockyduan, jonathanho, pabbeel}@eecs.berkeley.edu {marcin, bstadie, jonas, ilyasu,...
6709 |@word trial:2 kulis:1 seems:2 proportion:1 replicate:1 open:1 pieter:10 simulation:1 propagate:1 pick:3 versatile:1 shot:16 reduction:1 initial:3 liu:1 configuration:8 contains:1 jimenez:2 bc:3 o2:1 freitas:1 past:2 current:17 com:1 yuxuan:1 cad:1 activation:2 yet:1 diederik:1 must:2 guez:1 readily:2 john:3 devin...
6,312
671
An Information-Theoretic Approach to Deciphering the Hippocampal Code William E. Skaggs Bruce L. McNaughton Katalin M. Gothard Etan J. Markus Center for Neural Systems, Memory, and Aging 344 Life Sciences North University of Arizona Tucson AZ 85724 bill@nsma.arizona.edu Abstract Information theory is used to deriv...
671 |@word trial:2 briefly:1 hippocampus:3 seems:1 methodologically:1 pick:1 thereby:1 moment:1 configuration:1 series:1 interestingly:1 optican:2 current:1 nt:1 dx:2 must:2 distant:1 shape:3 plot:5 designed:2 drop:2 stationary:1 half:6 alone:1 cue:24 short:4 farther:1 provides:1 location:27 mathematical:1 constructed:...
6,313
6,710
Learning the Morphology of Brain Signals Using Alpha-Stable Convolutional Sparse Coding Mainak Jas1 , Tom Dupr? La Tour1 , Umut Sim? ? sekli1 , Alexandre Gramfort1,2 1: LTCI, Telecom ParisTech, Universit? Paris-Saclay, Paris, France 2: INRIA, Universit? Paris-Saclay, Saclay, France Abstract Neural time-series data co...
6710 |@word mild:1 trial:17 version:1 inversion:1 middle:1 norm:5 simulation:2 decomposition:2 contraction:1 eng:2 thereby:1 moment:1 liu:1 series:8 contains:1 document:2 deconvolutional:1 existing:1 current:2 recovered:1 activation:13 yet:3 written:2 must:1 csc:68 realistic:1 additive:3 plasticity:1 shape:3 mstep:1 mo...
6,314
6,711
Integration Methods and Optimization Algorithms Vincent Roulet INRIA, ENS, PSL Research University, Paris France vincent.roulet@inria.fr Damien Scieur INRIA, ENS, PSL Research University, Paris France damien.scieur@inria.fr Francis Bach INRIA, ENS, PSL Research University, Paris France francis.bach@inria.fr Alexandr...
6711 |@word version:1 polynomial:11 norm:1 disk:2 seek:3 simulation:1 tat:1 concise:1 initial:5 contains:1 loc:5 interestingly:1 past:1 kx0:1 discretization:4 comparing:1 written:3 must:1 numerical:15 designed:1 xk:54 oldest:1 beginning:1 recherche:1 provides:1 iterates:1 simpler:1 mathematical:2 c2:2 differential:11 i...
6,315
6,712
Sharpness, Restart and Acceleration Vincent Roulet INRIA, ENS Paris France vincent.roulet@inria.fr Alexandre d?Aspremont CNRS, ENS Paris France aspremon@ens.fr Abstract The ?ojasiewicz inequality shows that sharpness bounds on the minimum of convex optimization problems hold almost generically. Sharpness directly co...
6712 |@word repository:1 dtk:1 version:1 polynomial:2 norm:2 open:2 termination:2 seek:1 crucially:1 linearized:1 recursively:1 initial:4 juditski:3 si:5 bierstone:2 must:1 numerical:1 analytic:2 plot:1 une:1 xk:15 ojasiewicz:11 recherche:1 iterates:1 math:1 revisited:1 rc:1 along:1 mathematical:7 differential:1 initia...
6,316
6,713
Learning Koopman Invariant Subspaces for Dynamic Mode Decomposition Naoya Takeishi? , Yoshinobu Kawahara?,? , Takehisa Yairi? Department of Aeronautics and Astronautics, The University of Tokyo ? The Institute of Scientific and Industrial Research, Osaka University ? RIKEN Center for Advanced Intelligence Project {tak...
6713 |@word polynomial:3 norm:1 nd:3 casdagli:1 vogt:1 open:2 d2:1 r:17 decomposition:31 excited:1 sgd:6 mention:1 reduction:1 initial:3 liu:1 series:24 united:2 tuned:1 rightmost:2 past:1 existing:2 current:1 yairi:3 activation:3 must:6 numerical:8 plot:10 interpretable:1 update:3 mackey:1 intelligence:2 selected:1 ge...
6,317
6,714
Soft-to-Hard Vector Quantization for End-to-End Learning Compressible Representations Eirikur Agustsson ETH Zurich Fabian Mentzer ETH Zurich Michael Tschannen ETH Zurich aeirikur@vision.ee.ethz.ch mentzerf@vision.ee.ethz.ch michaelt@nari.ee.ethz.ch Lukas Cavigelli ETH Zurich Radu Timofte ETH Zurich & Merantix ...
6714 |@word pw:4 compression:55 replicate:1 retraining:1 open:1 simplifying:1 q1:1 sgd:2 thereby:1 solid:1 cleary:1 harder:1 initial:1 electronics:1 contains:1 daniel:2 denoting:1 document:2 ours:6 kurt:1 outperforms:1 current:2 laparra:2 discretization:2 activation:1 assigning:1 diederik:1 guez:1 attracted:1 john:3 re...
6,318
6,715
Learning spatiotemporal piecewise-geodesic trajectories from longitudinal manifold-valued data Juliette Chevallier CMAP, ?cole polytechnique juliette.chevallier@polytechnique.edu Pr St?phane Oudard Oncology Department USPC, AP-HP, HEGP St?phanie Allassonni?re CRC, Universit? Paris Descartes stephanie.allassonniere@p...
6715 |@word trial:2 version:5 seems:3 nd:2 open:2 iki:1 simulation:2 dominique:1 covariance:2 t1r:3 tr:33 solid:2 moment:1 initial:5 series:1 score:13 oldenburg:2 initialisation:2 longitudinal:7 past:1 yet:1 must:3 written:1 interrupted:1 numerical:3 additive:1 j1:15 realistic:1 shape:1 periodically:1 enables:1 asympto...
6,319
6,716
Improving Regret Bounds for Combinatorial Semi-Bandits with Probabilistically Triggered Arms and Its Applications Qinshi Wang Princeton University Princeton, NJ 08544 qinshiw@princeton.edu Wei Chen Microsoft Research Beijing, China weic@microsoft.com Abstract We study combinatorial multi-armed bandit with probabilis...
6716 |@word exploitation:2 briefly:2 version:6 eliminating:1 norm:14 stronger:2 hu:1 r:10 decomposition:1 lakshmanan:3 harder:2 liu:1 selecting:2 ours:1 prefix:1 past:1 existing:2 yajun:1 current:3 com:1 comparing:6 michal:1 si:5 conjunctive:7 maniu:1 happen:1 kdd:2 remove:6 update:1 greedy:2 selected:4 leaf:1 item:2 i...
6,320
6,717
Predictive-State Decoders: Encoding the Future into Recurrent Networks Arun Venkatraman1? , Nicholas Rhinehart1?, Wen Sun1 , Lerrel Pinto , Martial Hebert1 , Byron Boots2 , Kris M. Kitani1 , J. Andrew Bagnell1 1 The Robotics Institute, Carnegie-Mellon University, Pittsburgh, PA 2 School of Interactive Computing, Georg...
6717 |@word multitask:1 instrumental:2 kokkinos:1 nd:1 bptt:4 open:1 pieter:9 seek:1 propagate:1 simulation:1 r:1 psim:1 recursively:1 reduction:2 moment:4 configuration:1 series:3 daniel:2 lqr:1 precluding:1 rkhs:1 past:3 existing:4 current:5 yet:1 liva:1 must:1 john:4 devin:1 supervises:1 analytic:2 enables:1 hypothe...
6,321
6,718
Optimistic posterior sampling for reinforcement learning: worst-case regret bounds Shipra Agrawal Columbia University sa3305@columbia.edu Randy Jia Columbia University rqj2000@columbia.edu Abstract We present an algorithm based on posterior sampling (aka Thompson sampling) that achieves near-optimal worst-case regre...
6718 |@word h:1 exploitation:2 version:5 polynomial:1 stronger:2 nd:1 open:1 r:8 q1:1 pick:1 initial:3 liu:2 daniel:2 denoting:1 past:1 current:1 contextual:2 john:3 ronald:1 v:2 stationary:1 intelligence:2 beginning:5 ith:1 provides:2 unbounded:1 ucrl2:5 along:2 constructed:1 mathematical:3 katehakis:2 symposium:1 apo...
6,322
6,719
Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results Antti Tarvainen The Curious AI Company tarvaina@cai.fi Harri Valpola The Curious AI Company harri@cai.fi Abstract The recently proposed Temporal Ensembling has achieved state-of-the-art results in s...
6719 |@word cnn:4 middle:2 version:9 compression:3 seems:1 hu:1 rgb:1 bachman:2 sajjadi:1 accommodate:1 initial:1 liu:1 contains:1 uncovered:1 hoiem:1 daniel:1 ours:1 steiner:1 current:2 optim:1 activation:1 diederik:2 devin:1 additive:1 periodically:1 realistic:1 ronan:1 enables:1 wanted:1 christian:1 drop:1 update:1 ...
6,323
672
Object-Based Analog VLSI Vision Circuits Christof Koch Computation and Neural Systems California Institute of Technology Pasadena, CA Bimal Mathur, Shih-Chii Liu Rockwell International Science Center Thousand Oaks, CA John G. Harris MIT Artificial Intelligence Laboratory Cambridge, MA Jin Luo, Massimo Sivilotti Tann...
672 |@word seal:3 open:3 thereby:1 liu:7 contains:1 series:3 current:7 luo:5 yet:1 must:1 john:1 plot:2 v:1 intelligence:1 half:1 selected:1 compo:1 provides:2 node:8 location:2 oak:1 five:3 along:4 supply:1 consists:1 resistive:4 inside:8 behavior:1 frequently:1 globally:1 automatically:1 little:1 becomes:3 circuit:12...
6,324
6,720
Matching neural paths: transfer from recognition to correspondence search Nikolay Savinov1 1 Lubor Ladicky1 Marc Pollefeys1,2 Department of Computer Science at ETH Zurich, 2 Microsoft {nikolay.savinov,lubor.ladicky,marc.pollefeys}@inf.ethz.ch Abstract Many machine learning tasks require finding per-part correspon...
6720 |@word cnn:7 version:1 repository:1 polynomial:3 open:1 choy:1 p0:4 dramatic:1 mention:1 initial:1 contains:2 score:3 tuned:1 ours:15 prefix:3 rightmost:2 existing:1 com:1 activation:21 must:1 written:1 john:1 timestamps:2 visible:1 generative:1 colored:1 recompute:1 completeness:3 node:9 provides:1 ron:1 simpler:...
6,325
6,721
Linearly constrained Gaussian processes Carl Jidling Department of Information Technology Uppsala University, Sweden carl.jidling@it.uu.se Niklas Wahlstr?m Department of Information Technology Uppsala University, Sweden niklas.wahlstrom@it.uu.se Adrian Wills School of Engineering University of Newcastle, Australia a...
6721 |@word illustrating:2 version:2 norm:1 adrian:2 simulation:1 covariance:37 pick:1 contains:1 efficacy:1 salzmann:1 discretization:1 erms:3 ida:1 intriguing:1 written:5 must:3 john:2 numerical:4 shape:2 remove:2 plot:2 intelligence:2 selected:1 inspection:1 xk:6 short:1 provides:1 math:2 revisited:1 uppsala:3 gx:36...
6,326
6,722
Fixed-Rank Approximation of a Positive-Semidefinite Matrix from Streaming Data Joel A. Tropp Caltech Alp Yurtsever EPFL Madeleine Udell Cornell Volkan Cevher EPFL jtropp@caltech.edu alp.yurtsever@epfl.ch mru8@cornell.edu volkan.cevher@epfl.ch Abstract Several important applications, such as streaming PCA and semid...
6722 |@word trial:2 version:2 polynomial:2 norm:17 hu:1 confirms:1 covariance:4 decomposition:6 mention:1 nystr:33 solid:2 reduction:2 substitution:1 series:4 contains:3 selecting:1 woodruff:8 document:1 pprox:1 fa8750:1 existing:1 ketch:4 ka:15 chazelle:1 com:1 dx:1 must:5 fn:29 numerical:11 informative:2 remove:2 des...
6,327
6,723
Multi-Modal Imitation Learning from Unstructured Demonstrations using Generative Adversarial Nets Karol Hausman?? , Yevgen Chebotar??? , Stefan Schaal?? , Gaurav Sukhatme? , Joseph J. Lim? ? University of Southern California, Los Angeles, CA, USA Max-Planck-Institute for Intelligent Systems, T?bingen, Germany {hausma...
6723 |@word multitask:1 c0:2 adrian:1 instruction:1 pieter:7 cha:1 simulation:2 seek:1 covariance:1 p0:4 shot:1 reduction:1 initial:6 series:1 daniel:1 ours:1 hyunsoo:1 com:2 readily:1 john:2 enables:1 christian:1 interpretable:2 update:1 generative:11 leaf:1 imitate:16 imitated:1 accordingly:2 beginning:1 blei:1 detec...
6,328
6,724
Learning to Inpaint for Image Compression Mohammad Haris Baig? Department of Computer Science Dartmouth College Hanover, NH Vladlen Koltun Intel Labs Santa Clara, CA Lorenzo Torresani Dartmouth College Hanover, NH Abstract We study the design of deep architectures for lossy image compression. We present two archite...
6724 |@word cnn:1 middle:1 compression:42 loading:1 kokkinos:1 r:16 propagate:3 thereby:2 inpainting:69 harder:3 reduction:6 initial:3 contains:1 rippel:2 deconvolutional:1 outperforms:2 existing:5 recovered:1 current:3 com:1 laparra:1 guadarrama:1 clara:1 yet:1 diederik:1 must:4 bd:1 subsequent:5 additive:2 pertinent:...
6,329
6,725
Adaptive Bayesian Sampling with Monte Carlo EM Anirban Roychowdhury, Srinivasan Parthasarathy Department of Computer Science and Engineering The Ohio State University roychowdhury.7@osu.edu, srini@cse.ohio-state.edu Abstract We present a novel technique for learning the mass matrices in samplers obtained from discret...
6725 |@word kulis:2 version:2 inversion:1 dalal:1 nd:1 suitably:1 d2:1 simulation:3 covariance:10 sgd:2 thereby:3 initial:1 contains:1 efficacy:1 selecting:1 hereafter:1 series:4 denoting:2 document:7 existing:8 diagonalized:1 current:1 discretization:2 si:2 written:4 numerical:1 remove:2 plot:1 update:24 progressively...
6,330
6,726
ADMM without a Fixed Penalty Parameter: Faster Convergence with New Adaptive Penalization Yi Xu? , Mingrui Liu? , Qihang Lin? , Tianbao Yang? Department of Computer Science, The University of Iowa, Iowa City, IA 52242, USA ? Department of Management Sciences, The University of Iowa, Iowa City, IA 52242, USA {yi-xu, mi...
6726 |@word mild:1 version:1 polynomial:1 norm:3 nd:1 open:2 semicontinuous:3 linearized:21 covariance:5 decomposition:1 acknowlegements:1 reduction:1 initial:13 liu:6 series:2 tuned:1 interestingly:1 comparing:3 luo:2 tackling:1 chu:1 attracted:1 numerical:1 shape:2 update:11 implying:1 intelligence:1 prohibitive:1 we...
6,331
6,727
Shape and Material from Sound Zhoutong Zhang MIT Jiajun Wu MIT Qiujia Li University of Cambridge Zhengjia Huang ShanghaiTech University Joshua B. Tenenbaum MIT William T. Freeman MIT, Google Research Abstract What can we infer from hearing an object falling onto the ground? Based on knowledge of the physical worl...
6727 |@word kohli:1 illustrating:1 middle:1 open:2 simulation:16 excited:1 pressure:3 pick:2 harder:1 initial:5 configuration:2 contains:1 score:1 daniel:1 ours:1 past:7 existing:4 brien:5 current:2 comparing:1 synthesizer:1 familiarized:1 mesh:1 realistic:3 partition:1 informative:2 ronan:1 shape:36 enables:1 treating...
6,332
6,728
Flexible statistical inference for mechanistic models of neural dynamics Jan-Matthis Lueckmann? 1 , Pedro J. Gon?alves? 1 , Giacomo Bassetto1 , Kaan ?cal1,2 , Marcel Nonnenmacher1 , Jakob H. Macke?1 1 research center caesar, an associate of the Max Planck Society, Bonn, Germany 2 Mathematical Institute, University of ...
6728 |@word middle:2 nonsensical:2 open:2 prangle:1 grey:1 simulation:39 lezaun:1 covariance:5 eng:1 pg:7 solid:1 carry:1 reduction:1 initial:1 series:4 genetic:3 current:9 comparing:1 com:1 recovered:1 yet:2 assigning:1 fn:1 numerical:2 partition:1 informative:3 physiol:1 shape:1 enables:2 designed:3 update:1 implying...
6,333
6,729
Online Prediction with Selfish Experts Tim Roughgarden Department of Computer Science Stanford University Stanford, CA 94305 tim@cs.stanford.edu Okke Schrijvers Department of Computer Science Stanford University Stanford, CA 94305 okkes@cs.stanford.edu Abstract We consider the problem of binary prediction with expert...
6729 |@word mild:3 private:2 version:4 stronger:3 seems:2 dekel:2 bf:1 open:2 adrian:1 seek:1 simulation:7 forecaster:1 jacob:1 pick:2 incurs:2 solid:1 harder:2 liu:2 inefficiency:1 score:1 united:1 punishes:1 past:2 outperforms:1 current:1 com:6 savage:2 comparing:1 collude:3 si:2 must:1 john:2 numerical:2 informative...
6,334
673
A Hybrid Linear/Nonlinear Approach to Channel Equalization Problems Wei-Tsih Lee John Pearson David Sarnoff Research Center CN5300 Princeton, NJ 08543 Abstract Channel equalization problem is an important problem in high-speed communications. The sequences of symbols transmitted are distorted by neighboring symbols...
673 |@word version:2 inversion:7 seems:1 pulse:1 tried:1 ld:1 reduction:1 series:2 current:1 recovered:1 written:4 must:1 john:1 numerical:1 partition:1 v:1 cult:1 ith:1 provides:3 node:13 qam:6 location:1 constructed:4 direct:2 prove:1 combine:2 multi:1 company:1 increasing:6 confused:7 developed:3 nj:1 perfonn:1 xd:1...
6,335
6,730
Tensor Biclustering Soheil Feizi Stanford University sfeizi@stanford.edu Hamid Javadi Stanford University hrhakim@stanford.edu David Tse Stanford University dntse@stanford.edu Abstract Consider a dataset where data is collected on multiple features of multiple individuals over multiple times. This type of data can ...
6730 |@word trial:1 version:3 polynomial:2 norm:1 km:1 integrative:1 simulation:1 simplifying:1 decomposition:5 covariance:3 p0:6 moment:1 reduction:1 selecting:2 daniel:3 outperforms:3 olded:5 recovered:1 z2:6 com:1 activation:1 written:3 realistic:1 numerical:1 j1:64 drop:1 kv1:2 update:1 generative:1 vanishing:2 pro...
6,336
6,731
DPSCREEN: Dynamic Personalized Screening Kartik Ahuja Electrical and Computer Engineering Department University of California, Los Angeles ahujak@ucla.edu William R. Zame Economics Department University of California, Los Angeles zame@econ.ucla.edu Mihaela van der Schaar Engineering Science Department, University of...
6731 |@word trial:1 cox:3 seems:2 sex:1 pancreatic:6 simulation:3 sensed:1 incurs:1 carry:3 reduction:4 initial:4 score:1 efficacy:1 ours:1 expositional:1 longitudinal:2 past:2 existing:4 current:12 incidence:4 mihaela:2 riskier:1 si:1 must:2 john:1 subsequent:1 chicago:1 informative:1 benign:1 tailoring:1 designed:1 u...
6,337
6,732
Learning Unknown Markov Decision Processes: A Thompson Sampling Approach Yi Ouyang University of California, Berkeley ouyangyi@berkeley.edu Mukul Gagrani University of Southern California mgagrani@usc.edu Ashutosh Nayyar University of Southern California ashutosn@usc.edu Rahul Jain University of Southern California...
6732 |@word h:7 exploitation:3 version:2 polynomial:2 seems:1 simulation:4 pick:1 initial:1 liu:1 rightmost:1 outperforms:4 existing:1 current:2 nt:6 must:1 belmont:1 additive:1 numerical:3 ashutosh:2 update:1 resampling:4 stationary:14 v:2 selected:1 beginning:6 ith:2 provides:1 mannor:1 allerton:1 ucrl2:8 katehakis:1...
6,338
6,733
Testing and Learning on Distributions with Symmetric Noise Invariance Ho Chung Leon Law Department of Statistics University Of Oxford hlaw@stats.ox.ac.uk Christopher Yau Centre for Computational Biology University of Birmingham c.yau@bham.ac.uk Dino Sejdinovic Department of Statistics University Of Oxford dino.sejdi...
6733 |@word repository:1 stronger:1 norm:3 proportion:1 consolider:1 k2hk:1 prangle:1 km:1 simulation:4 azimuthal:1 accounting:1 decomposition:1 zolt:2 pick:1 carry:1 uncovered:1 contains:3 series:2 lichman:1 daniel:1 denoting:1 rkhs:1 interestingly:1 comparing:1 surprising:1 activation:1 diederik:1 must:1 written:4 ad...
6,339
6,734
A Dirichlet Mixture Model of Hawkes Processes for Event Sequence Clustering Hongteng Xu? School of ECE Georgia Institute of Technology hongtengxu313@gmail.com Hongyuan Zha College of Computing Georgia Institute of Technology zha@cc.gatech.edu Abstract How to cluster event sequences generated via different point proc...
6734 |@word multitask:1 trial:10 version:1 achievable:1 proportion:1 c0:2 rajaraman:1 open:10 initial:1 liu:1 contains:6 series:11 score:5 document:1 outperforms:2 existing:6 freitas:1 current:2 com:2 discretization:1 luo:4 gmail:1 vere:1 determinantal:1 subsequent:1 happen:1 kdd:2 nonnegativeness:1 designed:2 update:8...
6,340
6,735
Deanonymization in the Bitcoin P2P Network Giulia Fanti and Pramod Viswanath Abstract Recent attacks on Bitcoin?s peer-to-peer (P2P) network demonstrated that its transaction-flooding protocols, which are used to ensure network consistency, may enable user deanonymization?the linkage of a user?s IP address with her ps...
6735 |@word trial:2 version:1 stronger:1 seems:1 simulation:9 reap:1 pick:1 recursively:1 contains:4 selecting:1 janson:1 o2:1 outperforms:1 savage:1 com:4 virus:1 culprit:2 yet:2 must:1 parsing:1 written:2 realize:1 gpu:1 timestamps:25 enables:1 v:3 leaf:2 selected:1 fewer:1 discovering:1 vanishing:1 core:3 kairouz:1 ...
6,341
6,736
Accelerated consensus via Min-Sum Splitting Patrick Rebeschini Department of Statistics University of Oxford patrick.rebeschini@stats.ox.ac.uk Sekhar Tatikonda Department of Electrical Engineering Yale University sekhar.tatikonda@yale.edu Abstract We apply the Min-Sum message-passing protocol to solve the consensus ...
6736 |@word version:3 polynomial:2 norm:5 johansson:3 seems:1 open:1 d2:7 km:2 r:3 hu:1 tat:1 automat:1 incurs:1 initial:9 liu:1 daniel:1 tuned:1 current:1 chu:1 written:3 john:2 numerical:1 designed:1 update:7 maxv:1 parametrization:2 iterates:1 coarse:1 node:28 math:1 successive:1 allerton:1 location:1 zhang:1 constr...
6,342
6,737
Generalized Linear Model Regression under Distance-to-set Penalties Jason Xu University of California, Los Angeles jqxu@ucla.edu Eric C. Chi North Carolina State University eric_chi@ncsu.edu Kenneth Lange University of California, Los Angeles klange@ucla.edu Abstract Estimation in generalized linear models (GLM) is ...
6737 |@word mild:1 trial:2 illustrating:1 briefly:1 version:1 inversion:1 norm:11 seems:2 instrumental:1 proportionality:1 d2:6 seek:3 carolina:1 simulation:1 covariance:1 decomposition:2 invoking:1 mention:1 tr:1 carry:1 reduction:3 initial:1 series:6 score:1 ours:1 rightmost:1 outperforms:1 current:2 comparing:1 marq...
6,343
6,738
Adaptive stimulus selection for optimizing neural population responses Benjamin R. Cowley1,2 , Ryan C. Williamson1,2,5 , Katerina Acar2,6 , Matthew A. Smith?,2,7 , Byron M. Yu?,2,3,4 1 Machine Learning Dept., 2 Center for Neural Basis of Cognition, 3 Dept. of Electrical and Computer Engineering, 4 Dept. of Biomedical ...
6738 |@word neurophysiology:4 proceeded:1 cnn:55 middle:10 version:2 trial:18 norm:19 simulation:2 r:11 uncovers:2 shading:1 initial:1 liu:1 contains:1 series:1 selecting:3 genetic:3 existing:2 ninit:11 current:4 com:1 nt:5 scatter:14 yet:1 must:1 multineuron:1 subsequent:1 shape:3 plot:2 update:2 medial:1 v:2 alone:1 ...
6,344
6,739
Nonbacktracking Bounds on the Influence in Independent Cascade Models 1 Emmanuel Abbe1 2 Sanjeev Kulkarni2 Eun Jee Lee1 Program in Applied and Computational Mathematics 2 The Department of Electrical Engineering Princeton University {eabbe, kulkarni, ejlee}@princeton.edu Abstract This paper develops upper and lower ...
6739 |@word trial:1 briefly:1 version:3 instrumental:1 open:15 simulation:22 pulse:1 p0:2 lakshmanan:1 harder:1 recursively:3 initial:6 contains:1 score:4 neeman:1 outperforms:3 virus:1 si:1 yet:1 cis:1 remove:1 v:2 spec:1 selected:1 greedy:1 beginning:2 iterates:1 provides:1 node:45 zhang:2 mathematical:2 along:1 beco...
6,345
674
Directional-Unit Boltzmann Machines Richard S. Zemel Computer Science Dept. University of Toronto Toronto, ONT M5S lA4 Christopher K. I. Williams Computer Science Dept. University of Toronto Toronto, ONT M5S lA4 Michael C. Mozer Computer Science Dept. University of Colorado Boulder, CO 80309-0430 Abstract We present...
674 |@word version:3 simulation:2 tr:1 initial:1 cyclic:1 configuration:5 contains:2 series:1 current:1 nowlan:1 must:1 numerical:1 update:2 discrimination:1 reciprocal:3 steepest:1 provides:2 complication:1 toronto:6 attack:1 arctan:1 along:2 differential:2 consists:2 combine:2 baldi:2 hermitian:3 manner:2 expected:6 ...
6,346
6,740
Learning with Feature Evolvable Streams Bo-Jian Hou Lijun Zhang Zhi-Hua Zhou National Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, 210023, China {houbj,zhanglj,zhouzh}@lamda.nju.edu.cn Abstract Learning with streaming data has attracted much attention during the past few years. Though mo...
6740 |@word norm:1 nd:1 d2:5 contains:2 series:1 selecting:2 pt0:1 tuned:1 past:2 existing:2 outperforms:2 current:2 recovered:16 ka:1 protection:4 cumulation:1 si:1 yet:2 attracted:2 must:1 hou:2 drop:1 update:11 rd2:6 v:3 fund:1 half:1 intelligence:4 beginning:3 core:2 short:2 record:1 boosting:3 location:1 simpler:1...
6,347
6,741
Online Convex Optimization with Stochastic Constraints Hao Yu, Michael J. Neely, Xiaohan Wei Department of Electrical Engineering, University of Southern California? {yuhao,mjneely,xiaohanw}@usc.edu Abstract This paper considers online convex optimization (OCO) with stochastic constraints, which generalizes Zinkevich?...
6741 |@word norm:1 open:1 d2:11 q1:1 sgd:1 boundedness:2 configuration:1 series:1 selecting:1 existing:1 current:2 com:1 qureshi:3 universality:1 router:3 must:2 john:2 refines:1 numerical:1 happen:1 wellbehaved:1 plot:1 update:5 warmuth:2 beginning:3 iso:1 manfred:2 infrastructure:2 provides:2 mannor:5 simpler:1 unbou...
6,348
6,742
Max-Margin Invariant Features from Transformed Unlabeled Data Dipan K. Pal, Ashwin A. Kannan?, Gautam Arakalgud?, Marios Savvides Department of Electrical and Computer Engineering Carnegie Mellon University Pittsburgh, PA 15213 {dipanp,aalapakk,garakalgud,marioss}@cmu.edu Abstract The study of representations invarian...
6742 |@word briefly:1 version:7 covariance:1 thereby:8 harder:1 shot:3 moment:3 efficacy:1 score:1 tuned:1 rkhs:11 ours:3 past:1 existing:1 outperforms:3 stemmed:1 yet:2 written:1 readily:1 girosi:1 discrimination:2 intelligence:1 plane:1 ith:1 core:2 provides:3 boosting:1 bijection:1 gautam:1 gx:15 zhang:1 along:3 con...
6,349
6,743
Regularized Modal Regression with Applications in Cognitive Impairment Prediction Xiaoqian Wang1 , Hong Chen1 , Weidong Cai2 , Dinggang Shen3 , Heng Huang1? Department of Electrical and Computer Engineering, University of Pittsburgh, USA 2 School of Information Technologies, University of Sydney, Australia 3 Departmen...
6743 |@word mild:3 repository:2 version:2 mri:3 polynomial:2 norm:6 hu:1 seitz:1 carolina:1 lobe:6 p0:2 configuration:1 liu:1 score:5 selecting:1 lichman:1 mmse:1 longitudinal:2 outperforms:1 existing:2 com:1 comparing:2 gmail:1 written:1 john:1 informative:4 designed:1 half:4 epanechnikov:5 mental:2 characterization:2...
6,350
6,744
Translation Synchronization via Truncated Least Squares Xiangru Huang? The University of Texas at Austin 2317 Speedway, Austin, 78712 xrhuang@cs.utexas.edu Zhenxiao Liang? Tsinghua University Beijing, China, 100084 liangzx14@mails.tsinghua.edu.cn Chandrajit Bajaj The University of Texas at Austin 2317 Speedway, Aust...
6744 |@word qthat:1 kondor:2 norm:2 seitz:1 scg:1 decomposition:2 pick:1 concise:1 initial:14 series:1 contains:3 score:4 bc:1 outperforms:1 existing:1 current:2 recovered:1 comparing:1 com:1 nt:1 si:4 tackling:1 written:1 cruz:1 shakespeare:1 shape:4 enables:1 greedy:1 half:2 selected:2 directory:1 xk:9 chile:1 provid...
6,351
6,745
From which world is your graph? Cheng Li College of William & Mary Felix M. F. Wong Independent Researcher? Zhenming Liu College of William & Mary Varun Kanade University of Oxford Abstract Discovering statistical structure from links is a fundamental problem in the analysis of social networks. Choosing a misspeci...
6745 |@word mild:1 briefly:1 polynomial:5 stronger:2 proportion:1 nd:1 glue:1 suitably:3 c0:9 adrian:1 closure:1 eng:1 decomposition:1 paid:1 incarnation:1 reduction:3 liu:2 contains:6 score:4 daniel:1 neeman:2 ours:2 document:1 existing:2 err:2 whp:2 surprising:1 lang:2 intriguing:1 follower:2 must:1 john:2 realistic:...
6,352
6,746
A New Alternating Direction Method for Linear Programming Sinong Wang Department of ECE The Ohio State University wang.7691@osu.edu Ness Shroff Department of ECE and CSE The Ohio State University shroff.11@osu.edu Abstract It is well known that, for a linear program (LP) with constraint matrix A ? Rm?n , the Alterna...
6746 |@word mild:1 version:1 advantageous:1 norm:7 seems:1 open:3 simulation:1 linearized:1 tat:1 covariance:4 p0:1 hsieh:1 pick:1 d0k:2 tianyi:1 ipm:3 initial:1 inefficiency:1 contains:1 daniel:1 existing:11 current:7 recovered:1 luo:1 readily:1 numerical:2 subsequent:1 plot:1 designed:1 update:12 amir:1 une:1 xk:32 i...
6,353
6,747
Regret Analysis for Continuous Dueling Bandit Wataru Kumagai Center for Advanced Intelligence Project RIKEN 1-4-1, Nihonbashi, Chuo, Tokyo 103-0027, Japan wataru.kumagai@riken.jp Abstract The dueling bandit is a learning framework wherein the feedback information in the learning process is restricted to a noisy compa...
6747 |@word version:3 polynomial:1 norm:3 stronger:1 logit:1 dekel:1 open:1 d2:3 citeseer:1 pick:1 tr:1 moment:1 reduction:2 contains:1 tuned:1 ours:1 past:5 current:1 dikin:2 must:1 bd:3 numerical:2 enables:1 update:1 intelligence:1 math:1 preference:4 zhang:2 along:1 constructed:1 symposium:1 prove:2 busa:2 introduce...
6,354
6,748
Best Response Regression Omer Ben-Porat Technion - Israel Institute of Technology Haifa 32000 Israel omerbp@campus.technion.ac.il Moshe Tennenholtz Technion - Israel Institute of Technology Haifa 32000 Israel moshet@ie.technion.ac.il Abstract In a regression task, a predictor is given a set of instances, along with ...
6748 |@word briefly:1 polynomial:4 proportion:3 seems:3 dekel:1 willing:1 simulation:3 pick:2 thereby:3 solid:2 recursively:2 ld:4 reduction:1 contains:3 score:2 series:2 chervonenkis:1 interestingly:1 rightmost:1 outperforms:1 current:2 z2:1 com:1 intriguing:1 realistic:1 partition:1 additive:3 designed:1 maxv:1 intel...
6,355
6,749
TernGrad: Ternary Gradients to Reduce Communication in Distributed Deep Learning Wei Wen1 , Cong Xu2 , Feng Yan3 , Chunpeng Wu1 , Yandan Wang4 , Yiran Chen1 , Hai Li1 1 Duke University, 2 Hewlett Packard Labs, 3 University of Nevada ? Reno, 4 University of Pittsburgh 1 {wei.wen, chunpeng.wu, yiran.chen, hai.li}@duke....
6749 |@word cnn:1 polynomial:3 seems:1 norm:2 stronger:4 compression:1 heuristically:1 cipar:1 brightness:1 sgd:33 thereby:1 mention:2 nsw:1 reduction:1 initial:1 liu:2 lightweight:1 daniel:1 tuned:1 bradley:1 current:2 com:2 comparing:2 activation:1 diederik:1 danny:1 gpu:5 devin:2 numerical:6 christian:3 drop:2 plot:...
6,356
675
How Oscillatory Neuronal Responses Reflect Bistability and Switching of the Hidden Assembly Dynamics K. Pawelzik, H.-V. Bauert, J. Deppisch, and T. Geisel Institut fur Theoretische Physik and SFB 185 Nichtlineare Dynamik Universitat Frankfurt, Robert-Mayer-Str. 8-10, D-6000 Frankfurt/M. 11, FRG ttemporary adress:CNS-P...
675 |@word physik:1 confirms:1 tried:1 activation:1 yet:1 shape:2 designed:1 nichtlineare:1 selected:1 provides:2 successive:1 burst:13 hopf:2 autocorrelation:2 manner:1 inter:2 indeed:1 multi:5 brain:1 weinheim:2 pawelzik:10 str:1 pf:12 increasing:1 unpredictable:1 underlying:6 kind:1 dynamik:1 monkey:1 gottingen:1 te...
6,357
6,750
Learning Affinity via Spatial Propagation Networks Sifei Liu UC Merced, NVIDIA Guangyu Zhong Dalian University of Technology Shalini De Mello NVIDIA Ming-Hsuan Yang UC Merced, NVIDIA Jinwei Gu NVIDIA Jan Kautz NVIDIA Abstract In this paper, we propose spatial propagation networks for learning the affinity matrix fo...
6750 |@word cnn:27 middle:1 version:3 norm:2 kokkinos:1 everingham:1 paredes:1 seek:1 propagate:2 rgb:5 sgd:5 recursively:1 carry:1 initial:3 liu:7 configuration:2 contains:5 ndez:1 denoting:1 tuned:3 romera:1 existing:1 guadarrama:1 discretization:1 comparing:1 luo:2 must:1 parsing:8 shape:2 enables:2 designed:4 v:2 i...
6,358
6,751
Linear regression without correspondence Daniel Hsu Columbia University New York, NY djhsu@cs.columbia.edu Kevin Shi Columbia University New York, NY kshi@cs.columbia.edu Xiaorui Sun Microsoft Research Redmond, WA xiaoruisun@cs.columbia.edu Abstract This article considers algorithmic and statistical aspects of linea...
6751 |@word achievable:1 polynomial:5 norm:1 nd:1 c0:3 open:1 bn:1 decomposition:2 arjen:1 pick:2 reduction:13 initial:1 minw2rk:3 daniel:2 woodruff:1 recovered:1 z2:1 written:1 numerical:1 partition:3 klaas:1 generative:1 warmuth:3 beginning:1 short:1 manfred:2 completeness:1 quantized:3 allerton:1 simpler:1 mathemati...
6,359
6,752
NeuralFDR: Learning Discovery Thresholds from Hypothesis Features Fei Xia? , Martin J. Zhang?, James Zou? , David Tse? Stanford University {feixia,jinye,jamesz,dntse}@stanford.edu Abstract As datasets grow richer, an important challenge is to leverage the full features in the data to maximize the number of useful dis...
6752 |@word mild:1 version:1 proportion:23 stronger:3 c0:2 unif:3 willing:1 confirms:1 simulation:2 hu:2 accounting:1 ld:1 contains:4 score:1 series:2 genetic:3 ours:1 outperforms:2 bradley:1 com:1 wd:2 assigning:1 must:2 olive:1 john:2 partition:1 informative:2 shape:1 enables:2 pertinent:1 designed:1 interpretable:4 ...
6,360
6,753
Cost efficient gradient boosting Sven Peter Ferran Diego Heidelberg Collaboratory for Image Processing Interdisciplinary Center for Scientific Computing University of Heidelberg 69115 Heidelberg, Germany Robert Bosch GmbH Robert-Bosch-Stra?e 200 31139 Hildesheim, Germany ferran.diegoandilla@de.bosch.com sven.pete...
6753 |@word repository:1 briefly:2 compression:2 nd:2 confirms:1 pavel:1 citeseer:1 incurs:1 delgado:1 reduction:3 electronics:1 ndez:1 cristina:1 score:1 lichman:1 liu:1 daniel:2 document:3 greedymiser:2 outperforms:3 current:6 com:3 ka:1 manuel:1 activation:1 yet:4 john:3 ronald:1 subsequent:2 kdd:1 cheap:22 minmin:2...
6,361
6,754
Probabilistic Rule Realization and Selection Haizi Yu? ? Department of Computer Science University of Illinois at Urbana-Champaign Urbana, IL 61801 haiziyu7@illinois.edu Tianxi Li? Department of Statistics University of Michigan Ann Arbor, MI 48109 tianxili@umich.edu Lav R. Varshney? Department of Electrical and Comp...
6754 |@word middle:1 inversion:1 pw:5 norm:14 open:1 termination:2 linearized:1 perpin:1 decomposition:1 q1:1 pick:3 incurs:1 asks:1 mention:1 shot:1 moment:2 necessity:1 reduction:10 initial:1 score:1 selecting:3 interestingly:1 existing:2 err:3 ka:2 current:1 luo:1 lang:1 yet:3 tackling:1 chu:1 finest:1 exposing:1 re...
6,362
6,755
Nearest-Neighbor Sample Compression: Efficiency, Consistency, Infinite Dimensions Aryeh Kontorovich Department of Computer Science Ben-Gurion University of the Negev karyeh@cs.bgu.ac.il Sivan Sabato Department of Computer Science Ben-Gurion University of the Negev sabatos@bgu.ac.il Roi Weiss Department of Computer S...
6755 |@word h:2 version:2 compression:54 norm:1 yi0:4 open:10 decomposition:1 yjd:1 thereby:1 solid:1 carry:2 reduction:1 shechtman:1 celebrated:1 series:5 substitution:1 daniel:2 tuned:1 prefix:1 err:18 comparing:1 z2:2 surprising:4 ddim:5 beygelzimer:1 yet:2 intriguing:2 must:2 readily:1 john:4 michal:1 numerical:1 p...
6,363
6,756
A Scale Free Algorithm for Stochastic Bandits with Bounded Kurtosis Tor Lattimore? tor.lattimore@gmail.com Abstract Existing strategies for finite-armed stochastic bandits mostly depend on a parameter of scale that must be known in advance. Sometimes this is in the form of a bound on the payoffs, or the knowledge of a...
6756 |@word polynomial:1 achievable:2 seems:4 open:2 grey:1 calculus:1 specialises:1 boundedness:1 moment:6 united:1 kurt:4 existing:1 com:1 nt:1 gmail:1 must:1 written:1 partition:1 alone:1 ith:1 location:2 honda:6 mathematical:1 along:1 c2:18 katehakis:11 symposium:1 apostolos:1 indeed:1 expected:1 roughly:1 abbrevia...
6,364
6,757
Learning Multiple Tasks with Multilinear Relationship Networks Mingsheng Long, Zhangjie Cao, Jianmin Wang, Philip S. Yu School of Software, Tsinghua University, Beijing 100084, China {mingsheng,jimwang}@tsinghua.edu.cn caozhangjie14@gmail.com psyu@uic.edu Abstract Deep networks trained on large-scale data can learn...
6757 |@word multitask:6 determinant:1 cnn:7 stronger:1 paredes:2 chakraborty:1 open:2 d2:24 confirms:2 underperform:1 covariance:48 decomposition:4 jacob:1 sgd:1 versatile:1 tnlist:1 liu:1 contains:3 efficacy:3 selecting:2 tuned:1 suppressing:1 romera:2 outperforms:2 existing:2 com:2 transferability:8 nt:4 comparing:1 ...
6,365
6,758
Deep Hyperalignment Muhammad Yousefnezhad, Daoqiang Zhang College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics {myousefnezhad,dqzhang}@nuaa.edu.cn Abstract This paper proposes Deep Hyperalignment (DHA) as a regularized, deep extension, scalable Hyperalignment (HA) method, whic...
6758 |@word briefly:1 version:1 mri:1 norm:1 open:2 seek:6 simulation:1 r:1 decomposition:4 sgd:3 tr:5 reduction:1 series:2 contains:5 halchenko:1 bc:1 current:3 activation:2 tackling:1 must:7 gpu:1 shape:1 haxby:5 update:3 openfmri:1 fund:1 intelligence:1 selected:3 advancement:1 ubuntu:1 provides:4 location:1 allerto...
6,366
6,759
Online to Offline Conversions, Universality and Adaptive Minibatch Sizes Kfir Y. Levy Department of Computer Science, ETH Z?rich. yehuda.levy@inf.ethz.ch Abstract We present an approach towards convex optimization that relies on a novel scheme which converts adaptive online algorithms into offline methods. In the offl...
6759 |@word version:2 norm:15 dekel:1 open:1 d2:6 seek:1 invoking:4 sgd:14 harder:1 reduction:1 woodruff:1 interestingly:1 past:2 imaginary:4 universality:9 tackling:1 yet:1 written:1 john:1 intriguing:1 diederik:1 numerical:1 zaid:1 depict:1 update:13 v:1 juditsky:1 half:1 core:1 provides:2 zhang:2 mathematical:2 alon...
6,367
676
Destabilization and Route to Chaos in Neural Networks with Random Connectivity Bernard Doyon Unite INSERM 230 Service de Neurologie CHUPurpan F-31059 Toulouse Cedex, France Bruno Cessac Centre d'Etudes et de Recherches de Toulouse 2, avenue Edouard Belin, BP 4025 F-31055 Toulouse Cedex, France Mathias Quoy Centre d'...
676 |@word proportion:2 seems:1 disk:2 simulation:3 initial:1 born:1 series:1 ecole:1 manuel:1 activation:1 numerical:2 plot:1 stationary:3 selected:1 nervous:1 plane:1 math:2 sigmoidal:1 simpler:1 hopf:12 qualitative:1 cray:1 sustained:1 introduce:1 theoretically:1 expected:1 indeed:1 behavior:6 nor:1 brain:4 freeman:...
6,368
6,760
Stochastic Optimization with Variance Reduction for Infinite Datasets with Finite Sum Structure Alberto Bietti Inria? alberto.bietti@inria.fr Julien Mairal Inria? julien.mairal@inria.fr Abstract Stochastic optimization algorithms with variance reduction have proven successful for minimizing large finite sums of funct...
6760 |@word msr:1 version:1 middle:1 norm:3 seems:1 retraining:1 c0:11 open:1 tried:1 brightness:2 sgd:40 thereby:1 reduction:15 initial:6 contains:2 series:1 selecting:1 genetic:1 ours:2 document:1 tuned:1 outperforms:1 past:1 comparing:1 com:1 yet:1 tackling:1 written:1 tot:12 additive:2 ckns:1 realistic:1 shape:1 ho...
6,369
6,761
Deep Learning with Topological Signatures Roland Kwitt Department of Computer Science University of Salzburg, Austria Roland.Kwitt@sbg.ac.at Christoph Hofer Department of Computer Science University of Salzburg, Austria chofer@cosy.sbg.ac.at Marc Niethammer UNC Chapel Hill, NC, USA mn@cs.unc.edu Andreas Uhl Departme...
6761 |@word briefly:1 c0:4 decomposition:1 p0:2 homomorphism:2 kutzkov:1 sgd:2 bai:2 liu:2 series:1 contains:1 fragment:1 initial:2 ours:4 interestingly:2 outperforms:1 existing:3 steiner:3 current:1 discretization:1 z2:6 com:2 nanda:1 readily:1 realize:1 mesh:1 kdd:1 shape:18 enables:3 designed:1 fund:1 maxv:1 wasserm...
6,370
6,762
Predicting User Activity Level In Point Processes With Mass Transport Equation Yichen Wang? , Xiaojing Ye? , Hongyuan Zha? , Le Song? ? College of Computing, Georgia Institute of Technology ? School of Mathematics, Georgia State University {yichen.wang}@gatech.edu, xye@gsu.edu {zha,lsong}@cc.gatech.edu Abstract Point ...
6762 |@word multitask:1 seems:1 proportion:6 rajaraman:1 calculus:3 simulation:2 reduction:3 nonexistent:1 substitution:1 contains:9 liu:1 initial:7 past:2 existing:2 outperforms:1 unction:1 follower:1 numerical:1 informative:1 kdd:1 shape:2 enables:1 designed:3 plot:3 update:1 v:9 half:2 advancement:1 item:19 paramete...
6,371
6,763
Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues Noga Alon Tel Aviv University, Israel and Microsoft Research nogaa@tau.ac.il Moshe Babaioff Microsoft Research moshe@microsoft.com Yannai A. Gonczarowski The Hebrew University of Jerusalem, Israel and Microsoft Research yannai@gonch.name Shay Mo...
6763 |@word private:2 briefly:1 polynomial:3 stronger:1 willing:2 jacob:1 attainable:1 pick:1 thereby:1 boundedness:3 moment:17 contains:2 series:1 united:1 chervonenkis:2 interestingly:1 envision:2 ironing:1 com:3 gmail:2 fn:13 additive:6 confirming:1 n0:19 item:3 amir:3 yannai:4 short:1 core:1 alexandros:1 characteri...
6,372
6,764
Deep Dynamic Poisson Factorization Model Chengyue Gong Department of Information Management Peking University cygong@pku.edu.cn Win-bin Huang Department of Information Management Peking University huangwb@pku.edu.cn Abstract A new model, named as deep dynamic poisson factorization model, is proposed in this paper fo...
6764 |@word loading:2 liu:1 series:2 score:1 contains:4 united:5 africa:2 com:2 activation:1 written:1 lauly:1 additive:1 kdd:1 shape:1 interpretable:1 update:3 intelligence:1 fewer:1 greedy:1 generative:1 short:1 core:1 blei:5 five:5 along:1 augmentable:1 beta:1 prove:1 wale:1 fitting:5 combine:1 icews:10 bility:1 gro...
6,373
6,765
Positive-Unlabeled Learning with Non-Negative Risk Estimator Ryuichi Kiryo1,2 Gang Niu1,2 Marthinus C. du Plessis Masashi Sugiyama2,1 1 The University of Tokyo, 7-3-1 Hongo, Tokyo 113-0033, Japan 2 RIKEN, 1-4-1 Nihonbashi, Tokyo 103-0027, Japan { kiryo@ms., gang@ms., sugi@ }k.u-tokyo.ac.jp Abstract From only positive...
6765 |@word mild:1 kgk:2 cnn:1 norm:2 advantageous:1 stronger:1 bpu:23 softsign:4 proportion:1 open:1 tried:1 bn:2 contraction:1 rgb:1 p0:18 hsieh:1 sgd:1 solid:1 reduction:4 liu:5 series:1 seriously:1 document:3 existing:2 current:3 com:3 activation:1 yet:2 lang:1 must:2 written:2 gpu:9 realize:1 john:1 partition:1 kd...
6,374
6,766
Optimal Sample Complexity of M -wise Data for Top-K Ranking Minje Jang? School of Electrical Engineering KAIST jmj427@kaist.ac.kr Sunghyun Kim? Electronics and Telecommunications Research Institute Daejeon, Korea koishkim@etri.re.kr Changho Suh School of Electrical Engineering KAIST chsuh@kaist.ac.kr Sewoong Oh Indu...
6766 |@word trial:4 version:3 instrumental:1 norm:3 logit:1 c0:2 heuristically:2 simulation:5 pick:1 solid:1 recursively:1 moment:1 reduction:3 electronics:1 series:1 score:13 janson:2 horvitz:1 bradley:3 si:9 yet:2 numerical:8 partition:2 happen:1 drop:1 alone:3 stationary:2 item:47 beginning:1 reciprocal:1 vanishing:...
6,375
6,767
Reliable Decision Support using Counterfactual Models Suchi Saria Department of Computer Science Johns Hopkins University Baltimore, MD 21211 ssaria@cs.jhu.edu Peter Schulam Department of Computer Science Johns Hopkins University Baltimore, MD 21211 pschulam@cs.jhu.edu Abstract Making a good decision involves conside...
6767 |@word multitask:1 trial:1 longterm:1 polynomial:1 johansson:2 nd:1 grey:3 additively:1 simulation:1 hu:1 accounting:1 decomposition:1 covariance:7 creatinine:14 solid:1 initial:3 series:10 contains:2 score:14 ours:1 interestingly:1 longitudinal:3 past:1 reaction:1 horvitz:1 com:1 nt:10 vere:3 must:4 john:2 writte...
6,376
6,768
QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding Dan Alistarh IST Austria & ETH Zurich dan.alistarh@ist.ac.at Demjan Grubic ETH Zurich & Google demjangrubic@gmail.com Ryota Tomioka Microsoft Research ryoto@microsoft.com Jerry Z. Li MIT jerryzli@mit.edu Milan Vojnovic London School of Econom...
6768 |@word private:1 faculty:1 msr:1 compression:9 norm:1 version:9 proportion:1 bekkerman:1 briefly:1 open:2 achievable:1 bn:1 sgd:42 tr:1 solid:1 recursively:1 reduction:10 initial:3 liu:2 moment:10 contains:3 daniel:1 tuned:1 denoting:1 kurt:1 prefix:1 past:2 outperforms:1 current:5 com:4 luo:1 yet:1 gmail:1 must:2...
6,377
6,769
Convergent Block Coordinate Descent for Training Tikhonov Regularized Deep Neural Networks Ziming Zhang and Matthew Brand Mitsubishi Electric Research Laboratories (MERL) Cambridge, MA 02139-1955 {zzhang, brand}@merl.com Abstract By lifting the ReLU function into a higher dimensional space, we develop a smooth multi-...
6769 |@word trial:1 polynomial:1 norm:3 seems:2 nd:2 reused:1 mitsubishi:1 propagate:1 decomposition:1 sgd:16 arous:1 initial:3 contains:1 ours:1 guadarrama:1 com:2 comparing:1 luo:1 activation:6 dx:2 chu:1 gpu:1 numerical:1 oberman:1 update:2 v:4 stationary:8 half:1 lky:2 parameterization:1 accordingly:3 xk:7 vanishin...
6,378
677
A Formal Model of the Insect Olfactory Macroglomerulus: Simulations and Analytical Results. Christiane Linster David Marsan ESPCI, Laboratoire d'Electronique 10, Rue Vauquelin 75005 Paris, France Claudine Masson Laboratoire de Neurobiologie Comparee des Invertebrees INRA/CNRS (URA 1190) 91140 Bures sur Yvette, France ...
677 |@word middle:1 sex:2 simulation:17 lobe:3 excited:2 mammal:1 reaction:1 john:2 physiol:1 designed:2 discrimination:2 nq:1 recherche:1 compo:2 direct:1 differential:2 consists:1 pathway:2 olfactory:20 indeed:1 behavior:3 p1:1 brain:2 pf:1 ua:1 project:1 bounded:1 transformation:1 oscillates:1 classifier:2 unit:2 gr...
6,379
6,770
Train longer, generalize better: closing the generalization gap in large batch training of neural networks Elad Hoffer1?, Itay Hubara?, Daniel Soudry2 1 Technion - Israel Institute of Technology, Haifa, Israel 2 Columbia University, New York, New York, USA {elad.hoffer, itayhubara, daniel.soudry}@gmail.com Abstract Ba...
6770 |@word briefly:1 seems:2 norm:7 open:2 heuristically:1 bn:7 covariance:6 decomposition:1 incurs:1 sgd:18 arous:1 initial:13 daniel:2 tuned:1 document:1 interestingly:1 current:4 com:2 comparing:1 activation:1 gmail:1 yet:6 attracted:1 must:1 scatter:1 realistic:1 numerical:1 subsequent:1 partition:1 shape:1 enable...
6,380
6,771
Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks Urs K?ster* , Tristan Webb* , Xin Wang* , Marcel Nassar* , Arjun Bansal, William Constable, Oguz Elibol, Stewart Hall, Luke Hornof, Amir Khosrowshahi, Carey Kloss, Ruby Pai, Naveen Rao Artificial Intelligence Products Group, Intel Co...
6771 |@word trial:1 version:1 eliminating:2 annapureddy:1 tensorial:1 open:1 reused:1 simulation:1 accommodate:1 shading:1 pub:1 daniel:1 existing:2 bitwise:1 current:1 com:2 activation:14 yet:1 written:1 gpu:5 must:2 readily:1 numerical:18 periodically:1 additive:1 enables:1 designed:5 drop:1 update:10 yinda:1 preempt...
6,381
6,772
Model evidence from nonequilibrium simulations Michael Habeck Statistical Inverse Problems in Biophysics, Max Planck Institute for Biophysical Chemistry & Institute for Mathematical Stochastics, University of G?ttingen, 37077 G?ttingen, Germany email mhabeck@gwdg.de Abstract The marginal likelihood, or model evidence...
6772 |@word version:3 seems:3 confirms:1 simulation:54 p0:5 contrastive:2 pick:1 minus:1 carry:1 initial:5 configuration:3 contains:1 liu:1 document:1 ka:1 com:1 jaynes:1 yet:1 dx:2 realistic:1 partition:2 kpf:1 visible:3 heir:1 update:1 stationary:6 generative:1 selected:1 guess:1 intelligence:2 xk:16 geyer:1 core:1 s...
6,382
6,773
Minimal Exploration in Structured Stochastic Bandits Richard Combes Centrale-Supelec / L2S richard.combes@supelec.fr Stefan Magureanu KTH, EE School / ACL magur@kth.se Alexandre Proutiere KTH, EE School / ACL alepro@kth.se Abstract This paper introduces and addresses a wide class of stochastic bandit problems where ...
6773 |@word trial:1 exploitation:6 version:1 briefly:1 simplifying:1 attainable:1 pick:1 necessity:1 selecting:1 outperforms:2 existing:5 comparing:1 optim:1 contextual:1 yet:3 must:3 numerical:5 update:1 intelligence:1 selected:8 metrika:1 recherche:1 colored:1 mannor:2 complication:1 revisited:1 honda:2 math:1 simple...
6,383
6,774
Learned D-AMP: Principled Neural Network based Compressive Image Recovery Christopher A. Metzler Rice University chris.metzler@rice.edu Ali Mousavi Rice University ali.mousavi@rice.edu Richard G. Baraniuk Rice University richb@rice.edu Abstract Compressive image recovery is a challenging problem that requires fast ...
6774 |@word cnn:1 briefly:1 version:1 mri:3 compression:2 norm:1 blu:1 d2:2 seek:1 sensed:1 born:1 contains:2 selecting:1 mag:2 tuned:1 amp:43 mmse:5 outperforms:2 existing:1 comparing:1 discretization:2 com:1 yet:1 must:4 gpu:1 axk22:1 realistic:1 additive:1 enables:2 remove:1 designed:10 interpretable:2 update:2 poly...
6,384
6,775
Deliberation Networks: Sequence Generation Beyond One-Pass Decoding ? 1 Yingce Xia, 2 Fei Tian, 3 Lijun Wu, 1 Jianxin Lin, 2 Tao Qin, 1 Nenghai Yu, 2 Tie-Yan Liu 1 University of Science and Technology of China, Hefei, China 2 3 Microsoft Research, Beijing, China Sun Yat-sen University, Guangzhou, China 1 yingce.xia@g...
6775 |@word cnn:1 briefly:1 middle:7 seems:1 d2:14 paid:1 sgd:2 carry:1 initial:1 liu:10 series:1 score:11 contains:2 configuration:1 document:1 subword:1 past:1 outperforms:3 com:3 contextual:5 gmail:1 parmar:1 gpu:1 refines:4 concatenate:1 nian:1 remove:1 update:1 generative:1 selected:1 intelligence:2 krikun:1 short...
6,385
6,776
Adaptive Clustering through Semidefinite Programming Martin Royer Laboratoire de Math?matiques d?Orsay, Univ. Paris-Sud, CNRS, Universit? Paris-Saclay, 91405 Orsay, France martin.royer@math.u-psud.fr Abstract We analyze the clustering problem through a flexible probabilistic model that aims to identify an optimal par...
6776 |@word version:4 polynomial:1 achievable:2 norm:7 c0:2 open:1 simulation:2 bn:2 tr:9 reduction:1 configuration:1 score:1 mixon:1 recovered:1 comparing:2 written:1 numerical:2 partition:12 isotropic:4 detecting:1 math:2 node:1 provides:1 c6:2 c2:2 shorthand:1 introduce:4 peng:2 homoscedasticity:1 indeed:3 roughly:1...
6,386
6,777
Log-normality and Skewness of Estimated State/Action Values in Reinforcement Learning Liangpeng Zhang1,2 , Ke Tang3,1 , and Xin Yao3,2 1 School of Computer Science and Technology, University of Science and Technology of China 2 University of Birmingham, U.K. 3 Shenzhen Key Lab of Computational Intelligence, Department...
6777 |@word illustrating:1 stronger:1 smirnov:1 nd:2 d2:1 pieter:1 citeseer:1 solid:1 recursively:1 moment:1 initial:1 selecting:1 daniel:1 hasselt:3 current:2 comparing:1 analysed:1 si:24 yet:1 guez:2 john:3 timestamps:1 happen:1 informative:1 realistic:1 predetermined:1 cheap:1 update:1 overshooting:1 stationary:1 in...
6,387
6,778
Repeated Inverse Reinforcement Learning Kareem Amin? Google Research New York, NY 10011 kamin@google.com Nan Jiang? Satinder Singh Computer Science & Engineering, University of Michigan, Ann Arbor, MI 48104 {nanjiang,baveja}@umich.edu Abstract We introduce a novel repeated Inverse Reinforcement Learning problem: the...
6778 |@word private:1 version:4 polynomial:1 stronger:2 norm:1 unif:1 d2:3 pieter:4 jacob:1 deems:1 reduction:1 initial:13 inefficiency:1 contains:2 score:2 daniel:1 existing:1 err:1 current:1 com:1 comparing:1 yet:1 must:1 john:2 realize:1 ronald:1 subsequent:1 additive:1 remove:1 update:6 unidentifiability:5 greedy:1...
6,388
6,779
The Numerics of GANs Lars Mescheder Autonomous Vision Group MPI T?bingen lars.mescheder@tuebingen.mpg.de Sebastian Nowozin Machine Intelligence and Perception Group Microsoft Research sebastian.nowozin@microsoft.com Andreas Geiger Autonomous Vision Group MPI T?bingen andreas.geiger@tuebingen.mpg.de Abstract In this ...
6779 |@word norm:2 open:3 p0:4 nsw:3 thereby:1 inpainting:1 score:3 selecting:1 imaginary:11 current:2 com:2 contextual:1 john:2 devin:1 realistic:1 numerical:3 informative:1 enables:1 christian:1 update:1 depict:2 stationary:8 intelligence:1 generative:14 half:1 alec:2 accordingly:1 plane:1 iterates:2 pascanu:1 revisi...
6,389
678
Physiologically Based Speech Synthesis ~akoto Hirayanaa t ATR Human Information Processing Research Laboratories 2-2, Hikaridai, Seika-cho, Soraku-gun, Kyoto 619-02 Japan Eric Vatikiotis-Bateson tATR Auditory and Visual Perception Research Laboratories Kiyoshi Hondat Yasuharu Koiket ~itsuo Kawatot* Abstract This ...
678 |@word kura:1 blade:1 initial:1 anterior:1 wakita:2 synthesizer:1 yet:2 toh:1 numerical:1 subsequent:2 shape:5 motor:13 medial:1 v:1 device:1 yoh:1 record:1 honda:3 along:1 skilled:1 direct:1 emma:1 introduce:2 acquired:2 ra:1 behavior:6 seika:1 torque:1 encouraging:1 window:1 considering:1 increasing:1 begin:1 pro...
6,390
6,780
Practical Bayesian Optimization for Model Fitting with Bayesian Adaptive Direct Search Luigi Acerbi? Center for Neural Science New York University luigi.acerbi@nyu.edu Wei Ji Ma Center for Neural Science & Dept. of Psychology New York University weijima@nyu.edu Abstract Computational models in fields such as computa...
6780 |@word exploitation:1 version:5 middle:5 briefly:2 nd:4 mockus:1 zilinskas:1 calculus:1 simulation:5 seek:1 crucially:3 covariance:11 accounting:3 pick:1 initial:7 configuration:5 series:2 pub:3 sobol:2 tuned:1 genetic:2 interestingly:1 luigi:3 existing:1 ninit:2 freitas:3 current:12 outperforms:3 subjective:1 com...
6,391
6,781
Learning Chordal Markov Networks via Branch and Bound Kari Rantanen HIIT, Dept. Comp. Sci., University of Helsinki Antti Hyttinen HIIT, Dept. Comp. Sci., University of Helsinki Matti J?rvisalo HIIT, Dept. Comp. Sci., University of Helsinki Abstract We present a new algorithmic approach for the task of finding a chor...
6781 |@word middle:2 version:2 wiesel:1 polynomial:1 nd:2 giudici:1 heuristically:1 tried:1 bn:16 covariance:1 thereby:3 solid:1 versatile:1 initial:1 contains:4 score:22 past:1 current:10 comparing:1 chordal:28 john:1 tarantola:1 enables:1 update:2 fund:1 v:3 newest:1 intelligence:3 fewer:1 malone:2 selected:1 beginni...
6,392
6,782
Revenue Optimization with Approximate Bid Predictions ? Medina Andr?es Munoz Google Research 76 9th Ave New York, NY 10011 Sergei Vassilvitskii Google Research 76 9th Ave New York, NY 10011 Abstract In the context of advertising auctions, finding good reserve prices is a notoriously challenging learning problem. This...
6782 |@word version:1 achievable:2 polynomial:5 leighton:3 twelfth:1 willing:2 attainable:1 reduction:3 celebrated:1 contains:2 selecting:1 outperforms:2 current:1 surprising:2 si:3 sergei:2 must:2 readily:1 tenet:1 refines:1 adexchange:1 partition:23 shape:2 designed:1 plot:1 half:1 item:5 desktop:1 beginning:1 stest:...
6,393
6,783
Solving Most Systems of Random Quadratic Equations Gang Wang?,? ? Georgios B. Giannakis? Yousef Saad? Jie Chen? Key Lab of Intell. Contr. and Decision of Complex Syst., Beijing Inst. of Technology ? Digital Tech. Center & Dept. of Electrical and Computer Eng., Univ. of Minnesota ? Department of Computer Science an...
6783 |@word trial:4 phasemax:2 polynomial:1 norm:5 suitably:1 c0:2 confirms:1 rgb:1 eng:1 attainable:1 incurs:1 carry:1 shechtman:1 initial:5 series:1 efficacy:2 hereafter:1 score:2 mag:1 outperforms:1 existing:1 past:1 current:3 recovered:3 com:1 optim:1 tackling:1 yet:1 readily:2 additive:3 numerical:12 benign:1 pert...
6,394
6,784
Unsupervised Learning of Disentangled and Interpretable Representations from Sequential Data Wei-Ning Hsu, Yu Zhang, and James Glass Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge, MA 02139, USA {wnhsu,yzhang87,glass}@csail.mit.edu Abstract We present a factori...
6784 |@word kohli:1 middle:2 hu:1 pieter:1 grey:1 covariance:2 eng:1 configuration:3 contains:8 jimenez:2 outperforms:2 z2:81 com:1 lang:1 yet:1 diederik:4 john:2 distant:1 subsequent:2 designed:2 interpretable:10 drop:1 update:1 alone:1 intelligence:2 generative:17 alec:1 isotropic:2 fabius:1 short:5 regressive:2 prov...
6,395
6,785
Lookahead Bayesian Optimization with Inequality Constraints Remi R. Lam Massachusetts Institute of Technology Cambridge, MA rlam@mit.edu Karen E. Willcox Massachusetts Institute of Technology Cambridge, MA kwillcox@mit.edu Abstract We consider the task of optimizing an objective function subject to inequality constr...
6785 |@word exploitation:4 illustrating:1 seems:1 nd:1 mockus:1 zilinskas:1 seek:1 simulation:10 thereby:1 shading:1 recursively:1 reduction:5 initial:2 ndez:3 series:1 disparity:1 past:1 existing:1 current:2 com:1 surprising:1 yet:1 must:1 john:1 fn:1 numerical:6 subsequent:1 cheap:4 analytic:1 designed:2 update:1 dep...
6,396
6,786
Hierarchical Methods of Moments Matteo Ruffini ? Universitat Polit?cnica de Catalunya Guillaume Rabusseau ? McGill University Borja Balle ? Amazon Research Cambridge Abstract Spectral methods of moments provide a powerful tool for learning the parameters of latent variable models. Despite their theoretical appeal, t...
6786 |@word polynomial:2 stronger:1 proportion:1 open:1 d2:1 decomposition:13 recursively:3 moment:30 initial:1 liu:1 contains:3 series:1 daniel:6 genetic:1 document:11 outperforms:1 existing:6 diagonalized:1 recovered:2 com:1 comparing:2 written:1 john:1 numerical:2 partition:2 confirming:1 kdd:1 remove:2 designed:1 h...
6,397
6,787
Interpretable and Globally Optimal Prediction for Textual Grounding using Image Concepts Raymond A. Yeh, Jinjun Xiong? , Minh N. Do, Wen-mei W. Hwu, Alexander G. Schwing Department of Electrical Engineering, University of Illinois at Urbana-Champaign ? IBM Thomas J. Watson Research Center yeh17@illinois.edu, jinj...
6787 |@word kong:2 norm:1 kokkinos:1 underline:1 everingham:1 stronger:1 open:1 termination:1 hu:4 r:1 decomposition:4 mention:1 recursively:1 necessity:1 configuration:4 contains:2 score:26 cellphone:2 liu:1 initial:3 denoting:2 ours:6 tuned:3 fragment:1 outperforms:1 existing:3 current:1 com:1 guadarrama:2 contextual...
6,398
6,788
Revisit Fuzzy Neural Network: Demystifying Batch Normalization and ReLU with Generalized Hamming Network Lixin Fan lixin.fan@nokia.com Nokia Technologies Tampere, Finland Abstract We revisit fuzzy neural network with a cornerstone notion of generalized hamming distance, which provides a novel and theoretically justif...
6788 |@word kulis:1 cnn:4 middle:2 briefly:1 nd:1 open:1 closure:2 calculus:2 hu:1 bn:13 prominence:1 thres:4 lepetit:1 reduction:1 celebrated:3 liu:2 series:1 wj2:1 kurt:1 existing:1 current:1 com:3 manuel:1 analysed:1 activation:12 yet:2 intriguing:1 must:2 readily:2 john:1 luis:1 diederik:2 chu:1 confirming:1 christ...
6,399
6,789
Speeding Up Latent Variable Gaussian Graphical Model Estimation via Nonconvex Optimization Pan Xu Department of Computer Science University of Virginia Charlottesville, VA 22904 px3ds@virginia.edu Jian Ma School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213 jianma@cs.cmu.edu Quanquan Gu Departme...
6789 |@word determinant:7 briefly:1 inversion:2 norm:25 bf:3 c0:5 d2:7 confirms:1 r:5 covariance:11 decomposition:9 contraction:1 tr:4 initial:3 configuration:1 liu:9 score:3 tuned:2 ours:3 xinyang:1 nonparanormal:1 outperforms:1 existing:3 nicolai:1 luo:1 toh:1 yet:1 written:1 bd:3 john:3 numerical:2 designed:1 plot:1...