Jennifer G. Dy, Andreas Krause 0001
Proceedings of the 35th International Conference on Machine Learning
ICML, 2018.

@proceedings{ICML-2018,
	editor        = "Jennifer G. Dy and Andreas Krause 0001",
	ee            = "http://proceedings.mlr.press/v80/",
	publisher     = "{PMLR}",
	series        = "{Proceedings of Machine Learning Research}",
	title         = "{Proceedings of the 35th International Conference on Machine Learning}",
	volume        = 80,
	year          = 2018,
}

Contents (621 items)

ICML-2018-AbeilleL #bound #linear #polynomial #problem

Improved Regret Bounds for Thompson Sampling in Linear Quadratic Control Problems (MA, AL), pp. 1–9.

ICML-2018-AbelALL #abstraction #learning

State Abstractions for Lifelong Reinforcement Learning (DA, DA, LL, MLL), pp. 10–19.

ICML-2018-AbelJGKL #learning #policy

Policy and Value Transfer in Lifelong Reinforcement Learning (DA, YJ, SYG, GDK, MLL), pp. 20–29.

ICML-2018-AcharyaKSZ #named

INSPECTRE: Privately Estimating the Unseen (JA, GK0, ZS, HZ), pp. 30–39.

ICML-2018-AchlioptasDMG #3d #generative #learning #modelling

Learning Representations and Generative Models for 3D Point Clouds (PA, OD, IM, LJG), pp. 40–49.

ICML-2018-AdelGW #generative #modelling

Discovering Interpretable Representations for Both Deep Generative and Discriminative Models (TA, ZG, AW), pp. 50–59.

ICML-2018-AgarwalBD0W #approach #classification #reduction

A Reductions Approach to Fair Classification (AA, AB, MD, JL0, HMW), pp. 60–69.

ICML-2018-AgarwalPA #ranking

Accelerated Spectral Ranking (AA, PP, SA0), pp. 70–79.

ICML-2018-AghazadehSLDSB #feature model #named #scalability #sketching #using

MISSION: Ultra Large-Scale Feature Selection using Count-Sketches (AA, RS, DL, GD, AS, RGB), pp. 80–88.

ICML-2018-AgrawalUB #graph #modelling #scalability

Minimal I-MAP MCMC for Scalable Structure Discovery in Causal DAG Models (RA, CU, TB), pp. 89–98.

ICML-2018-0001ZM #distributed

Proportional Allocation: Simple, Distributed, and Diverse Matching with High Entropy (SA0, MZ, VSM), pp. 99–108.

ICML-2018-AhnCWS #approximate

Bucket Renormalization for Approximate Inference (SA, MC, AW, JS), pp. 109–118.

ICML-2018-AinsworthFLF #analysis #named

oi-VAE: Output Interpretable VAEs for Nonlinear Group Factor Analysis (SKA, NJF, AKCL, EBF), pp. 119–128.

ICML-2018-AlaaS #algorithm #design #guidelines

Limits of Estimating Heterogeneous Treatment Effects: Guidelines for Practical Algorithm Design (AMA, MvdS), pp. 129–138.

ICML-2018-AlaaS18a #automation #kernel #learning #modelling #named #optimisation

AutoPrognosis: Automated Clinical Prognostic Modeling via Bayesian Optimization with Structured Kernel Learning (AMA, MvdS), pp. 139–148.

ICML-2018-Alabdulmohsin #empirical #optimisation #scalability

Information Theoretic Guarantees for Empirical Risk Minimization with Applications to Model Selection and Large-Scale Optimization (IMA), pp. 149–158.

ICML-2018-AlemiPFDS0

Fixing a Broken ELBO (AAA, BP, IF, JVD, RAS, KM0), pp. 159–168.

ICML-2018-AliakbarpourDR #equivalence #testing

Differentially Private Identity and Equivalence Testing of Discrete Distributions (MA, ID, RR), pp. 169–178.

ICML-2018-Allen-Zhu #optimisation #probability

Katyusha X: Practical Momentum Method for Stochastic Sum-of-Nonconvex Optimization (ZAZ), pp. 179–185.

ICML-2018-Allen-ZhuBL #bound #first-order

Make the Minority Great Again: First-Order Regret Bound for Contextual Bandits (ZAZ, SB, YL), pp. 186–194.

ICML-2018-AlmahairiRSBC #learning

Augmented CycleGAN: Learning Many-to-Many Mappings from Unpaired Data (AA, SR, AS, PB, ACC), pp. 195–204.

ICML-2018-AmitM

Meta-Learning by Adjusting Priors Based on Extended PAC-Bayes Theory (RA, RM), pp. 205–214.

ICML-2018-AmodioK #biology #named

MAGAN: Aligning Biological Manifolds (MA, SK), pp. 215–223.

ICML-2018-AndoniLSZZ #linear

Subspace Embedding and Linear Regression with Orlicz Norm (AA, CL0, YS0, PZ, RZ), pp. 224–233.

ICML-2018-ArenzZN #performance #policy #using

Efficient Gradient-Free Variational Inference using Policy Search (OA, MZ, GN), pp. 234–243.

ICML-2018-AroraCH #network #on the #optimisation

On the Optimization of Deep Networks: Implicit Acceleration by Overparameterization (SA, NC, EH), pp. 244–253.

ICML-2018-Arora0NZ #approach #bound

Stronger Generalization Bounds for Deep Nets via a Compression Approach (SA, RG0, BN, YZ), pp. 254–263.

ICML-2018-AsadiML #learning #modelling

Lipschitz Continuity in Model-based Reinforcement Learning (KA, DM, MLL), pp. 264–273.

ICML-2018-AthalyeC0 #obfuscation #security

Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples (AA, NC, DAW0), pp. 274–283.

ICML-2018-AthalyeEIK #robust

Synthesizing Robust Adversarial Examples (AA, LE, AI, KK), pp. 284–293.

ICML-2018-AwasthiV #clustering

Clustering Semi-Random Mixtures of Gaussians (PA, AV), pp. 294–303.

ICML-2018-BacciuEM #approach #generative #graph #markov

Contextual Graph Markov Model: A Deep and Generative Approach to Graph Processing (DB, FE, AM), pp. 304–313.

ICML-2018-BaiB

Greed is Still Good: Maximizing Monotone Submodular+Supermodular (BP) Functions (WB, JAB), pp. 314–323.

ICML-2018-Baity-JesiSGSAC #network

Comparing Dynamics: Deep Neural Networks versus Glassy Systems (MBJ, LS, MG, SS, GBA, CC, YL, MW, GB), pp. 324–333.

ICML-2018-BajajGHHL #clustering #named #using

SMAC: Simultaneous Mapping and Clustering Using Spectral Decompositions (CB, TG, ZH, QH, ZL), pp. 334–343.

ICML-2018-BajgarKK #architecture #performance

A Boo(n) for Evaluating Architecture Performance (OB, RK, JK), pp. 344–352.

ICML-2018-BalcanDSV #branch #learning

Learning to Branch (MFB, TD, TS, EV), pp. 353–362.

ICML-2018-BalduzziRMFTG #game studies

The Mechanics of n-Player Differentiable Games (DB, SR, JM, JNF, KT, TG), pp. 363–372.

ICML-2018-BalestrieroCGB #learning

Spline Filters For End-to-End Deep Learning (RB, RC, HG, RGB), pp. 373–382.

ICML-2018-BalestrieroB #network

A Spline Theory of Deep Networks (RB, RGB), pp. 383–392.

ICML-2018-BalkanskiS #adaptation #approximate

Approximation Guarantees for Adaptive Sampling (EB, YS), pp. 393–402.

ICML-2018-BalleW #difference #privacy

Improving the Gaussian Mechanism for Differential Privacy: Analytical Calibration and Optimal Denoising (BB, YXW), pp. 403–412.

ICML-2018-BallesH #probability

Dissecting Adam: The Sign, Magnitude and Variance of Stochastic Gradients (LB, PH), pp. 413–422.

ICML-2018-BalogTS #database #kernel

Differentially Private Database Release via Kernel Mean Embeddings (MB, IOT, BS), pp. 423–431.

ICML-2018-BamlerM #modelling #optimisation #symmetry

Improving Optimization in Models With Continuous Symmetry Breaking (RB, SM), pp. 432–441.

ICML-2018-BangS #generative #network #using

Improved Training of Generative Adversarial Networks using Representative Features (DB, HS), pp. 442–451.

ICML-2018-BansalAB #agile #design #using

Using Inherent Structures to design Lean 2-layer RBMs (AB, AA, CB), pp. 452–460.

ICML-2018-BaoNS #classification #similarity

Classification from Pairwise Similarity and Unlabeled Data (HB, GN, MS), pp. 461–470.

ICML-2018-BaptistaP #combinator #optimisation

Bayesian Optimization of Combinatorial Structures (RB, MP), pp. 471–480.

ICML-2018-BaqueRFF #optimisation

Geodesic Convolutional Shape Optimization (PB, ER, FF, PF), pp. 481–490.

ICML-2018-BargiacchiVRNH #coordination #graph #learning #multi #problem

Learning to Coordinate with Coordination Graphs in Repeated Single-Stage Multi-Agent Decision Problems (EB, TV, DMR, AN, HvH), pp. 491–499.

ICML-2018-BarmanBG #testing

Testing Sparsity over Known and Unknown Bases (SB, AB0, SG), pp. 500–509.

ICML-2018-BarretoBQSSHMZM #learning #policy #using

Transfer in Deep Reinforcement Learning Using Successor Features and Generalised Policy Improvement (AB, DB, JQ, TS, DS, MH, DJM, AZ, RM), pp. 510–519.

ICML-2018-BartlettHL #linear

Gradient descent with identity initialization efficiently learns positive definite linear transformations (PLB, DPH, PML), pp. 520–529.

ICML-2018-BelghaziBROBHC #estimation

Mutual Information Neural Estimation (MIB, AB, SR, SO, YB, RDH, ACC), pp. 530–539.

ICML-2018-BelkinMM #kernel #learning

To Understand Deep Learning We Need to Understand Kernel Learning (MB, SM, SM), pp. 540–548.

ICML-2018-BenderKZVL #architecture #comprehension

Understanding and Simplifying One-Shot Architecture Search (GB, PJK, BZ, VV, QVL), pp. 549–558.

ICML-2018-BernsteinWAA #named #optimisation #problem

SIGNSGD: Compressed Optimisation for Non-Convex Problems (JB, YXW, KA, AA), pp. 559–568.

ICML-2018-BhaskaraW #clustering #distributed

Distributed Clustering via LSH Based Data Partitioning (AB, MW), pp. 569–578.

ICML-2018-BinkowskiMD #network

Autoregressive Convolutional Neural Networks for Asynchronous Time Series (MB, GM, PD), pp. 579–588.

ICML-2018-BlancR #adaptation #kernel

Adaptive Sampled Softmax with Kernel Based Sampling (GB, SR), pp. 589–598.

ICML-2018-BojanowskiJLS #generative #network #optimisation

Optimizing the Latent Space of Generative Networks (PB, AJ, DLP, AS), pp. 599–608.

ICML-2018-BojchevskiSZG #generative #graph #named #random

NetGAN: Generating Graphs via Random Walks (AB, OS, DZ, SG), pp. 609–618.

ICML-2018-BollapragadaMNS #machine learning

A Progressive Batching L-BFGS Method for Machine Learning (RB, DM, JN, HJMS, PTPT), pp. 619–628.

ICML-2018-BonakdarpourCBL #predict

Prediction Rule Reshaping (MB, SC, RFB, JL), pp. 629–637.

ICML-2018-BoracchiCCM #data type #detection #multi #named

QuantTree: Histograms for Change Detection in Multivariate Data Streams (GB, DC, CC, DM), pp. 638–647.

ICML-2018-BravermanCKLWY #data type #matrix #multi #performance

Matrix Norms in Data Streams: Faster, Multi-Pass and Row-Order (VB, SRC, RK, YL0, DPW, LFY), pp. 648–657.

ICML-2018-BrukhimG #modelling #predict

Predict and Constrain: Modeling Cardinality in Deep Structured Prediction (NB, AG), pp. 658–666.

ICML-2018-BuchholzWM #monte carlo

Quasi-Monte Carlo Variational Inference (AB, FW, SM), pp. 667–676.

ICML-2018-CaiYZHY #architecture #network #performance

Path-Level Network Transformation for Efficient Architecture Search (HC, JY, WZ0, SH, YY0), pp. 677–686.

ICML-2018-CalandrielloKLV #graph #learning #scalability

Improved Large-Scale Graph Learning through Ridge Spectral Sparsification (DC, IK, AL, MV), pp. 687–696.

ICML-2018-CampbellB

Bayesian Coreset Construction via Greedy Iterative Geodesic Ascent (TC, TB), pp. 697–705.

ICML-2018-CaoGWSHT #coordination #learning

Adversarial Learning with Local Coordinate Coding (JC, YG, QW, CS, JH, MT), pp. 706–714.

ICML-2018-CelisKS0KV #summary

Fair and Diverse DPP-Based Data Summarization (LEC, VK, DS, AD0, TK, NKV), pp. 715–724.

ICML-2018-CeylanG #estimation #modelling

Conditional Noise-Contrastive Estimation of Unnormalised Models (CC, MUG), pp. 725–733.

ICML-2018-ChapfuwaTLPGCH #modelling

Adversarial Time-to-Event Modeling (PC, CT, CL, CP, BG, LC, RH), pp. 734–743.

ICML-2018-CharlesP #algorithm #learning

Stability and Generalization of Learning Algorithms that Converge to Global Optima (ZBC, DSP), pp. 744–753.

ICML-2018-Chatterjee #learning

Learning and Memorization (SC), pp. 754–762.

ICML-2018-ChatterjiFMBJ #formal method #monte carlo #on the #probability #reduction

On the Theory of Variance Reduction for Stochastic Gradient Monte Carlo (NSC, NF, YAM, PLB, MIJ), pp. 763–772.

ICML-2018-ChatziafratisNC #clustering #constraints

Hierarchical Clustering with Structural Constraints (VC, RN, MC), pp. 773–782.

ICML-2018-ChePLJL #generative #modelling #multi

Hierarchical Deep Generative Models for Multi-Rate Multivariate Time Series (ZC, SP, MGL, BJ, YL0), pp. 783–792.

ICML-2018-ChenBLR #adaptation #multi #named #network #normalisation

GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks (ZC0, VB, CYL, AR), pp. 793–802.

ICML-2018-0003FK #constraints #question

Weakly Submodular Maximization Beyond Cardinality Constraints: Does Randomization Help Greedy? (LC0, MF, AK), pp. 803–812.

ICML-2018-ChenHHK #online #optimisation #probability

Projection-Free Online Optimization with Stochastic Gradient: From Convexity to Submodularity (LC0, CH, HH, AK), pp. 813–822.

ICML-2018-ChenLCWPC #estimation #performance

Continuous-Time Flows for Efficient Inference and Density Estimation (CC, CL, LC, WW, YP, LC), pp. 823–832.

ICML-2018-ChenLW #learning #scalability #using

Scalable Bilinear Learning Using State and Action Features (YC, LL0, MW), pp. 833–842.

ICML-2018-ChenMGBO

Stein Points (WYC, LWM, JG, FXB, CJO), pp. 843–852.

ICML-2018-ChenMS #learning

Learning K-way D-dimensional Discrete Codes for Compact Embedding Representations (TC0, MRM, YS), pp. 853–862.

ICML-2018-ChenMRA #generative #named

PixelSNAIL: An Improved Autoregressive Generative Model (XC0, NM, MR, PA), pp. 863–871.

ICML-2018-ChenPS #network

Dynamical Isometry and a Mean Field Theory of RNNs: Gating Enables Signal Propagation in Recurrent Neural Networks (MC, JP, SSS), pp. 872–881.

ICML-2018-ChenSWJ #learning

Learning to Explain: An Information-Theoretic Perspective on Model Interpretation (JC, LS, MJW, MIJ), pp. 882–891.

ICML-2018-ChenTZHC #bound

Variational Inference and Model Selection with Generalized Evidence Bounds (LC, CT, RZ, RH, LC), pp. 892–901.

ICML-2018-ChenWCP #distributed #named

DRACO: Byzantine-resilient Distributed Training via Redundant Gradients (LC, HW, ZBC, DSP), pp. 902–911.

ICML-2018-ChenXCY #adaptation #named #probability

SADAGRAD: Strongly Adaptive Stochastic Gradient Methods (ZC, YX, EC, TY), pp. 912–920.

ICML-2018-ChenXW0G #estimation #matrix #optimisation #precise

Covariate Adjusted Precision Matrix Estimation via Nonconvex Optimization (JC, PX0, LW, JM0, QG), pp. 921–930.

ICML-2018-ChenXG #learning #multi

End-to-End Learning for the Deep Multivariate Probit Model (DC, YX, CPG), pp. 931–940.

ICML-2018-ChenZS #graph #network #probability #reduction

Stochastic Training of Graph Convolutional Networks with Variance Reduction (JC, JZ0, LS), pp. 941–949.

ICML-2018-ChengDH #learning #rank

Extreme Learning to Rank via Low Rank Assumption (MC, ID, CJH), pp. 950–959.

ICML-2018-Chierichetti0T #learning #multi

Learning a Mixture of Two Multinomial Logits (FC, RK0, AT), pp. 960–968.

ICML-2018-ChoromanskiRSTW #architecture #evolution #optimisation #policy #scalability

Structured Evolution with Compact Architectures for Scalable Policy Optimization (KC, MR, VS, RET, AW), pp. 969–977.

ICML-2018-ChowNG #consistency #learning

Path Consistency Learning in Tsallis Entropy Regularized MDPs (YC, ON, MG), pp. 978–987.

ICML-2018-ChowdhuryYD #framework #sketching

An Iterative, Sketching-based Framework for Ridge Regression (AC, JY, PD), pp. 988–997.

ICML-2018-ClaiciCS #probability

Stochastic Wasserstein Barycenters (SC, EC, JS), pp. 998–1007.

ICML-2018-Co-ReyesLGEAL #learning #self

Self-Consistent Trajectory Autoencoder: Hierarchical Reinforcement Learning with Trajectory Embeddings (JDCR, YL, AG0, BE, PA, SL), pp. 1008–1017.

ICML-2018-CohenDO #on the

On Acceleration with Noise-Corrupted Gradients (MC0, JD, LO), pp. 1018–1027.

ICML-2018-CohenHKLMT #linear #online #polynomial

Online Linear Quadratic Control (AC, AH, TK, NL, YM, KT), pp. 1028–1037.

ICML-2018-ColasSO #algorithm #learning #named

GEP-PG: Decoupling Exploration and Exploitation in Deep Reinforcement Learning Algorithms (CC, OS, PYO), pp. 1038–1047.

ICML-2018-CormodeDW #distributed #streaming #summary

Leveraging Well-Conditioned Bases: Streaming and Distributed Summaries in Minkowski p-Norms (GC, CD, DPW), pp. 1048–1056.

ICML-2018-CorneilGB #learning #performance

Efficient ModelBased Deep Reinforcement Learning with Variational State Tabulation (DSC, WG, JB), pp. 1057–1066.

ICML-2018-CortesDGMY #learning #online

Online Learning with Abstention (CC, GD, CG, MM, SY), pp. 1067–1075.

ICML-2018-CotterFYGB

Constrained Interacting Submodular Groupings (AC, MMF, SY, MRG, JAB), pp. 1076–1085.

ICML-2018-CremerLD

Inference Suboptimality in Variational Autoencoders (CC, XL, DD), pp. 1086–1094.

ICML-2018-CzarneckiJJHTHO #education #learning

Mix & Match Agent Curricula for Reinforcement Learning (WMC, SMJ, MJ, LH, YWT, NH, SO, RP), pp. 1095–1103.

ICML-2018-DabneyOSM #learning #network

Implicit Quantile Networks for Distributional Reinforcement Learning (WD, GO, DS, RM), pp. 1104–1113.

ICML-2018-DaiKDSS #algorithm #graph #learning

Learning Steady-States of Iterative Algorithms over Graphs (HD, ZK, BD, AJS, LS), pp. 1114–1122.

ICML-2018-DaiLTHWZS #graph

Adversarial Attack on Graph Structured Data (HD, HL, TT0, XH, LW, JZ0, LS), pp. 1123–1132.

ICML-2018-DaiS0XHLCS #approximate #convergence #learning #named

SBEED: Convergent Reinforcement Learning with Nonlinear Function Approximation (BD, AS, LL0, LX, NH, ZL0, JC, LS), pp. 1133–1142.

ICML-2018-DaiZGW #network #using

Compressing Neural Networks using the Variational Information Bottleneck (BD, CZ, BG, DPW), pp. 1143–1152.

ICML-2018-DamaskinosMGPT #machine learning

Asynchronous Byzantine Machine Learning (the case of SGD) (GD, EMEM, RG, RP, MT), pp. 1153–1162.

ICML-2018-DaneshmandKLH #probability

Escaping Saddles with Stochastic Gradients (HD, JMK, AL, TH), pp. 1163–1172.

ICML-2018-SaCW #modelling #scalability #visual notation

Minibatch Gibbs Sampling on Large Graphical Models (CDS, VC, WW), pp. 1173–1181.

ICML-2018-DentonF #generative #probability #video

Stochastic Video Generation with a Learned Prior (ED, RF), pp. 1182–1191.

ICML-2018-DepewegHDU #composition #learning #nondeterminism #performance

Decomposition of Uncertainty in Bayesian Deep Learning for Efficient and Risk-sensitive Learning (SD, JMHL, FDV, SU), pp. 1192–1201.

ICML-2018-DeshpandeMST #adaptation #linear #modelling

Accurate Inference for Adaptive Linear Models (YD, LWM, VS, MT), pp. 1202–1211.

ICML-2018-DezfouliBN #network

Variational Network Inference: Strong and Stable with Concrete Support (AD, EVB, RN), pp. 1212–1221.

ICML-2018-DharGE #generative #modelling #using

Modeling Sparse Deviations for Compressed Sensing using Generative Models (MD, AG, SE), pp. 1222–1231.

ICML-2018-DiakonikolasO #coordination #random

Alternating Randomized Block Coordinate Descent (JD, LO), pp. 1232–1240.

ICML-2018-DibangoyeB #distributed #learning

Learning to Act in Decentralized Partially Observable MDPs (JSD, OB), pp. 1241–1250.

ICML-2018-DiengRAB #named #network

Noisin: Unbiased Regularization for Recurrent Neural Networks (ABD, RR, JA, DMB), pp. 1251–1260.

ICML-2018-DietterichTC #learning

Discovering and Removing Exogenous State Variables and Rewards for Reinforcement Learning (TGD, GT, ZC), pp. 1261–1269.

ICML-2018-DimakopoulouR #concurrent #coordination #learning

Coordinated Exploration in Concurrent Reinforcement Learning (MD, BVR), pp. 1270–1278.

ICML-2018-DoerrDSNSTT #modelling #probability

Probabilistic Recurrent State-Space Models (AD, CD, MS, DNT, SS, MT, ST), pp. 1279–1288.

ICML-2018-DoikovR #polynomial #random

Randomized Block Cubic Newton Method (ND, PR), pp. 1289–1297.

ICML-2018-DouikH #clustering #graph #matrix #optimisation #probability #rank

Low-Rank Riemannian Optimization on Positive Semidefinite Stochastic Matrices with Applications to Graph Clustering (AD, BH), pp. 1298–1307.

ICML-2018-DraxlerVSH #energy #network

Essentially No Barriers in Neural Network Energy Landscape (FD, KV, MS, FAH), pp. 1308–1317.

ICML-2018-Drutsa #algorithm #consistency

Weakly Consistent Optimal Pricing Algorithms in Repeated Posted-Price Auctions with Strategic Buyer (AD), pp. 1318–1327.

ICML-2018-DuL #network #on the #polynomial #power of

On the Power of Over-parametrization in Neural Networks with Quadratic Activation (SSD, JDL), pp. 1328–1337.

ICML-2018-DuLTSP

Gradient Descent Learns One-hidden-layer CNN: Don't be Afraid of Spurious Local Minima (SSD, JDL, YT, AS, BP), pp. 1338–1347.

ICML-2018-DubeyAPGE #game studies #video

Investigating Human Priors for Playing Video Games (RD, PA, DP, TG, AAE), pp. 1348–1356.

ICML-2018-DunnerLGBHJ #algorithm #distributed #higher-order #trust

A Distributed Second-Order Algorithm You Can Trust (CD, AL, MG, AB, TH, MJ), pp. 1357–1365.

ICML-2018-DvurechenskyGK #algorithm #complexity

Computational Optimal Transport: Complexity by Accelerated Gradient Descent Is Better Than by Sinkhorn's Algorithm (PED, AG, AK), pp. 1366–1375.

ICML-2018-Dziugaite0 #bound

Entropy-SGD optimizes the prior of a PAC-Bayes bound: Generalization properties of Entropy-SGD and data-dependent priors (GKD, DMR0), pp. 1376–1385.

ICML-2018-EfroniDSM #approach #learning

Beyond the One-Step Greedy Approach in Reinforcement Learning (YE, GD, BS, SM), pp. 1386–1395.

ICML-2018-EsfandiariLM #algorithm #composition #parallel #streaming

Parallel and Streaming Algorithms for K-Core Decomposition (HE, SL, VSM), pp. 1396–1405.

ICML-2018-EspeholtSMSMWDF #architecture #distributed #named #scalability

IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures (LE, HS, RM, KS, VM, TW, YD, VF, TH, ID, SL, KK), pp. 1406–1415.

ICML-2018-EvansN #process #scalability

Scalable Gaussian Processes with Grid-Structured Eigenfunctions (GP-GRIEF) (TWE, PBN), pp. 1416–1425.

ICML-2018-FalahatgarJOPR #learning #ranking

The Limits of Maxing, Ranking, and Preference Learning (MF, AJ, AO, VP, VR), pp. 1426–1435.

ICML-2018-FalknerKH #named #optimisation #performance #robust #scalability

BOHB: Robust and Efficient Hyperparameter Optimization at Scale (SF, AK, FH), pp. 1436–1445.

ICML-2018-FarajtabarCG #evaluation #robust

More Robust Doubly Robust Off-policy Evaluation (MF, YC, MG), pp. 1446–1455.

ICML-2018-FathonyBZZ #consistency #performance

Efficient and Consistent Adversarial Bipartite Matching (RF, SB, XZ, BDZ), pp. 1456–1465.

ICML-2018-Fazel0KM #convergence #linear #policy #polynomial

Global Convergence of Policy Gradient Methods for the Linear Quadratic Regulator (MF, RG0, SMK, MM), pp. 1466–1475.

ICML-2018-FazelniaP #named

CRVI: Convex Relaxation for Variational Inference (GF, JWP), pp. 1476–1484.

ICML-2018-FellowsCW #fourier #policy

Fourier Policy Gradients (MF, KC, SW), pp. 1485–1494.

ICML-2018-FengWCS #learning #multi #network #parametricity #using

Nonparametric variable importance using an augmented neural network with multi-task learning (JF, BDW, MC, NS), pp. 1495–1504.

ICML-2018-FilstroffLF #matrix

Closed-form Marginal Likelihood in Gamma-Poisson Matrix Factorization (LF, AL, CF), pp. 1505–1513.

ICML-2018-FlorensaHGA #automation #generative #learning

Automatic Goal Generation for Reinforcement Learning Agents (CF, DH, XG, PA), pp. 1514–1523.

ICML-2018-FoersterFARXW #infinity #monte carlo #named

DiCE: The Infinitely Differentiable Monte Carlo Estimator (JNF, GF, MAS, TR, EPX, SW), pp. 1524–1533.

ICML-2018-FosterADLS

Practical Contextual Bandits with Regression Oracles (DJF, AA, MD, HL, RES), pp. 1534–1543.

ICML-2018-FraccaroRZPEV #generative #memory management #modelling

Generative Temporal Models with Spatial Memory for Partially Observed Environments (MF, DJR, YZ, AP, SMAE, FV), pp. 1544–1553.

ICML-2018-FrancaRV

ADMM and Accelerated ADMM as Continuous Dynamical Systems (GF, DPR, RV), pp. 1554–1562.

ICML-2018-FranceschiFSGP #optimisation #programming

Bilevel Programming for Hyperparameter Optimization and Meta-Learning (LF, PF, SS, RG, MP), pp. 1563–1572.

ICML-2018-FruitPLO #learning #performance

Efficient Bias-Span-Constrained Exploration-Exploitation in Reinforcement Learning (RF, MP, AL, RO), pp. 1573–1581.

ICML-2018-FujimotoHM #approximate #fault

Addressing Function Approximation Error in Actor-Critic Methods (SF, HvH, DM), pp. 1582–1591.

ICML-2018-FujitaM #policy

Clipped Action Policy Gradient (YF0, SiM), pp. 1592–1601.

ICML-2018-FurlanelloLTIA #network

Born-Again Neural Networks (TF, ZCL, MT, LI, AA), pp. 1602–1611.

ICML-2018-Gaboardi0 #testing

Local Private Hypothesis Testing: Chi-Square Tests (MG, RR0), pp. 1612–1621.

ICML-2018-GanapathiramanS #induction #modelling #parametricity

Inductive Two-layer Modeling with Parametric Bregman Transfer (VG, ZS, XZ, YY), pp. 1622–1631.

ICML-2018-GaneaBH #learning

Hyperbolic Entailment Cones for Learning Hierarchical Embeddings (OEG, GB, TH), pp. 1632–1641.

ICML-2018-GanianKOS #algorithm #matrix #problem

Parameterized Algorithms for the Matrix Completion Problem (RG, IAK, SO, SS), pp. 1642–1651.

ICML-2018-GaninKBEV #image #learning #source code #using

Synthesizing Programs for Images using Reinforced Adversarial Learning (YG, TK, IB, SMAE, OV), pp. 1652–1661.

ICML-2018-GaoCL #named #network #optimisation

Spotlight: Optimizing Device Placement for Training Deep Neural Networks (YG, LC0, BL), pp. 1662–1670.

ICML-2018-GaoW #learning #network #parallel

Parallel Bayesian Network Structure Learning (TG, DW), pp. 1671–1680.

ICML-2018-GarciaCEd #learning #predict

Structured Output Learning with Abstention: Application to Accurate Opinion Prediction (AG0, CC, SE, FdB), pp. 1681–1689.

ICML-2018-GarneloRMRSSTRE #process

Conditional Neural Processes (MG, DR, CM, TR, DS, MS, YWT, DJR, SMAE), pp. 1690–1699.

ICML-2018-GengKPP #modelling #visual notation

Temporal Poisson Square Root Graphical Models (SG, ZK, PLP, DP), pp. 1700–1709.

ICML-2018-Georgogiannis #fault #learning #taxonomy

The Generalization Error of Dictionary Learning with Moreau Envelopes (AG), pp. 1710–1718.

ICML-2018-GhassamiSKB #design #empirical #learning

Budgeted Experiment Design for Causal Structure Learning (AG, SS, NK, EB), pp. 1719–1728.

ICML-2018-GhodsLGS #linear #retrieval

Linear Spectral Estimators and an Application to Phase Retrieval (RG, ASL, TG, CS), pp. 1729–1738.

ICML-2018-GhoshYD #learning #network

Structured Variational Learning of Bayesian Neural Networks with Horseshoe Priors (SG, JY, FDV), pp. 1739–1748.

ICML-2018-GhoshalH #learning #modelling #polynomial #predict

Learning Maximum-A-Posteriori Perturbation Models for Structured Prediction in Polynomial Time (AG, JH), pp. 1749–1757.

ICML-2018-GibsonG #modelling #robust #scalability

Robust and Scalable Models of Microbiome Dynamics (TEG, GKG), pp. 1758–1767.

ICML-2018-GilraG #learning #network

Non-Linear Motor Control by Local Learning in Spiking Neural Networks (AG, WG), pp. 1768–1777.

ICML-2018-GoelKM #learning

Learning One Convolutional Layer with Overlapping Patches (SG, ARK, RM), pp. 1778–1786.

ICML-2018-GreydanusKDF #comprehension #visualisation

Visualizing and Understanding Atari Agents (SG, AK, JD, AF), pp. 1787–1796.

ICML-2018-GroverAGBE #learning #multi #policy

Learning Policy Representations in Multiagent Systems (AG, MAS, JKG, YB, HE), pp. 1797–1806.

ICML-2018-GuHDH #algorithm #memory management #performance #probability

Faster Derivative-Free Stochastic Algorithm for Shared Memory Machines (BG, ZH, CD, HH), pp. 1807–1816.

ICML-2018-GuezWASVWMS #learning

Learning to Search with MCTSnets (AG, TW, IA, KS, OV, DW, RM, DS), pp. 1817–1826.

ICML-2018-GunasekarLSS #bias #geometry #optimisation

Characterizing Implicit Bias in Terms of Optimization Geometry (SG, JDL, DS, NS), pp. 1827–1836.

ICML-2018-0001KS #named #optimisation #probability

Shampoo: Preconditioned Stochastic Tensor Optimization (VG0, TK, YS), pp. 1837–1845.

ICML-2018-HaarnojaHAL #learning #policy

Latent Space Policies for Hierarchical Reinforcement Learning (TH, KH, PA, SL), pp. 1846–1855.

ICML-2018-HaarnojaZAL #learning #probability

Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor (TH, AZ, PA, SL), pp. 1856–1865.

ICML-2018-HaghiriGL #random

Comparison-Based Random Forests (SH, DG, UvL), pp. 1866–1875.

ICML-2018-HammN #learning #optimisation #performance

K-Beam Minimax: Efficient Optimization for Deep Adversarial Learning (JH, YKN), pp. 1876–1884.

ICML-2018-HanHZ #classification #estimation #multi #problem #scalability

Candidates vs. Noises Estimation for Large Multi-Class Classification Problem (LH, YH, TZ), pp. 1885–1894.

ICML-2018-HanL

Stein Variational Gradient Descent Without Gradient (JH0, QL0), pp. 1895–1903.

ICML-2018-YeZ0Z #modelling #semantics

Rectify Heterogeneous Models with Semantic Mapping (HJY, DCZ, YJ0, ZHZ), pp. 1904–1913.

ICML-2018-HartfordGLR #interactive #modelling #set

Deep Models of Interactions Across Sets (JSH, DRG, KLB, SR), pp. 1914–1923.

ICML-2018-HashemiSSALCKR #data access #learning #memory management

Learning Memory Access Patterns (MH, KS, JAS, GA, HL, JC, CK, PR), pp. 1924–1933.

ICML-2018-HashimotoSNL

Fairness Without Demographics in Repeated Loss Minimization (TBH, MS, HN, PL), pp. 1934–1943.

ICML-2018-Hebert-JohnsonK #multi #named

Multicalibration: Calibration for the (Computationally-Identifiable) Masses (ÚHJ, MPK, OR, GNR), pp. 1944–1953.

ICML-2018-HefnyM0SG #network #policy #predict

Recurrent Predictive State Policy Networks (AH, ZM, WS0, SSS, GJG), pp. 1954–1963.

ICML-2018-HeinonenYMIL #learning #modelling #process

Learning unknown ODE models with Gaussian processes (MH, CY, HM, JI, HL), pp. 1964–1973.

ICML-2018-HelfrichWY #network #orthogonal

Orthogonal Recurrent Neural Networks with Scaled Cayley Transform (KH, DW, QY0), pp. 1974–1983.

ICML-2018-HoPW #performance #robust

Fast Bellman Updates for Robust MDPs (CPH, MP, WW), pp. 1984–1993.

ICML-2018-HoffmanTPZISED #adaptation #named

CyCADA: Cycle-Consistent Adversarial Domain Adaptation (JH, ET, TP, JYZ, PI, KS, AAE, TD), pp. 1994–2003.

ICML-2018-HoltzenBM #abstraction #composition #probability #source code

Sound Abstraction and Decomposition of Probabilistic Programs (SH, GVdB, TDM), pp. 2004–2013.

ICML-2018-HongRL #algorithm #distributed #higher-order #network #optimisation

Gradient Primal-Dual Algorithm Converges to Second-Order Stationary Solution for Nonconvex Distributed Optimization Over Networks (MH, MR, JDL), pp. 2014–2023.

ICML-2018-HronMG

Variational Bayesian dropout: pitfalls and fixes (JH, AGdGM, ZG), pp. 2024–2033.

ICML-2018-HuNSS #classification #learning #question #robust

Does Distributionally Robust Supervised Learning Give Robust Classifiers? (WH, GN, IS, MS), pp. 2034–2042.

ICML-2018-HuWL #analysis #probability #reduction #source code #using

Dissipativity Theory for Accelerating Stochastic Variance Reduction: A Unified Analysis of SVRG and Katyusha Using Semidefinite Programs (BH, SW, LL), pp. 2043–2052.

ICML-2018-Huang #matrix #sketching

Near Optimal Frequent Directions for Sketching Dense and Sparse Matrices (ZH), pp. 2053–2062.

ICML-2018-HuangA0S #learning #using

Learning Deep ResNet Blocks Sequentially using Boosting Theory (FH, JTA, JL0, RES), pp. 2063–2072.

ICML-2018-Huang0S #learning #markov #modelling #topic

Learning Hidden Markov Models from Pairwise Co-occurrences with Application to Topic Modeling (KH, XF0, NDS), pp. 2073–2082.

ICML-2018-HuangKLC

Neural Autoregressive Flows (CWH, DK, AL, ACC), pp. 2083–2092.

ICML-2018-Huntsman #estimation

Topological Mixture Estimation (SH), pp. 2093–2102.

ICML-2018-HuoGYH #convergence #parallel

Decoupled Parallel Backpropagation with Convergence Guarantee (ZH, BG, QY, HH), pp. 2103–2111.

ICML-2018-IcarteKVM #composition #learning #specification #using

Using Reward Machines for High-Level Task Specification and Decomposition in Reinforcement Learning (RTI, TQK, RAV, SAM), pp. 2112–2121.

ICML-2018-IglZLWW #learning

Deep Variational Reinforcement Learning for POMDPs (MI, LMZ, TAL, FW, SW), pp. 2122–2131.

ICML-2018-IlseTW #learning #multi

Attention-based Deep Multiple Instance Learning (MI, JMT, MW), pp. 2132–2141.

ICML-2018-IlyasEAL #black box #query

Black-box Adversarial Attacks with Limited Queries and Information (AI, LE, AA, JL), pp. 2142–2151.

ICML-2018-ImamuraSS #analysis #clustering #crowdsourcing #fault

Analysis of Minimax Error Rate for Crowdsourcing and Its Application to Worker Clustering Model (HI, IS, MS), pp. 2152–2161.

ICML-2018-ImaniW #performance

Improving Regression Performance with Distributional Losses (EI, MW), pp. 2162–2171.

ICML-2018-InouyeR

Deep Density Destructors (DII, PR), pp. 2172–2180.

ICML-2018-ItoYF #estimation #optimisation #predict

Unbiased Objective Estimation in Predictive Optimization (SI, AY, RF), pp. 2181–2190.

ICML-2018-IvanovB

Anonymous Walk Embeddings (SI, EB), pp. 2191–2200.

ICML-2018-JaffeWCKN #approach #learning #modelling

Learning Binary Latent Variable Models: A Tensor Eigenpair Approach (AJ, RW, SC, YK, BN), pp. 2201–2210.

ICML-2018-JainJ #optimisation

Firing Bandits: Optimizing Crowdfunding (LJ, KGJ), pp. 2211–2219.

ICML-2018-0002TT #matrix #revisited

Differentially Private Matrix Completion Revisited (PJ0, ODT, AT), pp. 2220–2229.

ICML-2018-JangKS #predict #video

Video Prediction with Appearance and Motion Conditions (YJ, GK, YS), pp. 2230–2239.

ICML-2018-JankowiakO

Pathwise Derivatives Beyond the Reparameterization Trick (MJ, FO), pp. 2240–2249.

ICML-2018-JanzingS #detection #linear #modelling #multi

Detecting non-causal artifacts in multivariate linear regression models (DJ, BS), pp. 2250–2258.

ICML-2018-JawanpuriaM #framework #learning #matrix #rank

A Unified Framework for Structured Low-rank Matrix Learning (PJ, BM), pp. 2259–2268.

ICML-2018-JeongS #learning #performance

Efficient end-to-end learning for quantizable representations (YJ, HOS), pp. 2269–2278.

ICML-2018-JiaLQA #network

Exploring Hidden Dimensions in Parallelizing Convolutional Neural Networks (ZJ, SL, CRQ, AA), pp. 2279–2288.

ICML-2018-JiangEL #learning

Feedback-Based Tree Search for Reinforcement Learning (DRJ, EE, HL), pp. 2289–2298.

ICML-2018-JiangJK

Quickshift++: Provably Good Initializations for Sample-Based Mean Shift (HJ, JJ, SK), pp. 2299–2308.

ICML-2018-JiangZLLF #data-driven #education #learning #named #network

MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels (LJ0, ZZ, TL, LJL, LFF0), pp. 2309–2318.

ICML-2018-JiaoV #higher-order #kernel #permutation

The Weighted Kendall and High-order Kernels for Permutations (YJ, JPV), pp. 2319–2327.

ICML-2018-JinBJ #generative #graph

Junction Tree Variational Autoencoder for Molecular Graph Generation (WJ, RB, TSJ), pp. 2328–2337.

ICML-2018-JinKL #network #testing

Network Global Testing by Counting Graphlets (JJ, ZTK, SL), pp. 2338–2346.

ICML-2018-JinKL18a #learning

Regret Minimization for Partially Observable Deep Reinforcement Learning (PHJ, KK, SL), pp. 2347–2356.

ICML-2018-JinYXYJFY #named #network #performance

WSNet: Compact and Efficient Networks Through Weight Sampling (XJ, YY, NX0, JY, NJ, JF, SY), pp. 2357–2366.

ICML-2018-JohnH #fourier #process #scalability #using

Large-Scale Cox Process Inference using Variational Fourier Features (STJ, JH), pp. 2367–2375.

ICML-2018-Johnson0 #functional #generative #learning #modelling

Composite Functional Gradient Learning of Generative Adversarial Models (RJ, TZ0), pp. 2376–2384.

ICML-2018-JoseCF

Kronecker Recurrent Units (CJ, MC, FF), pp. 2385–2394.

ICML-2018-KaiserBRVPUS #modelling #performance #sequence #using

Fast Decoding in Sequence Models Using Discrete Latent Variables (LK, SB, AR, AV, NP, JU, NS), pp. 2395–2404.

ICML-2018-KajiharaKYF #estimation #kernel #recursion

Kernel Recursive ABC: Point Estimation with Intractable Likelihood (TK, MK, KY, KF), pp. 2405–2414.

ICML-2018-KalchbrennerESN #performance #synthesis

Efficient Neural Audio Synthesis (NK, EE, KS, SN, NC, EL, FS, AvdO, SD, KK), pp. 2415–2424.

ICML-2018-KalimerisSSW #learning #using

Learning Diffusion using Hyperparameters (DK, YS, KS, UW), pp. 2425–2433.

ICML-2018-KallummilK #orthogonal #statistics

Signal and Noise Statistics Oblivious Orthogonal Matching Pursuit (SK, SK), pp. 2434–2443.

ICML-2018-KallusZ #machine learning

Residual Unfairness in Fair Machine Learning from Prejudiced Data (NK, AZ), pp. 2444–2453.

ICML-2018-KalyanLKB #learning #multi

Learn from Your Neighbor: Learning Multi-modal Mappings from Sparse Annotations (AK, SL, AK, DB), pp. 2454–2463.

ICML-2018-KamnitsasCFWTRG #clustering #learning

Semi-Supervised Learning via Compact Latent Space Clustering (KK, DCC, LLF, IW, RT, DR, BG, AC, AVN), pp. 2464–2473.

ICML-2018-KangJF #optimisation #policy

Policy Optimization with Demonstrations (BK, ZJ, JF), pp. 2474–2483.

ICML-2018-KangP #random

Improving Sign Random Projections With Additional Information (KK, WWP), pp. 2484–2492.

ICML-2018-KangarshahiHSC #framework #game studies

Let's be Honest: An Optimal No-Regret Framework for Zero-Sum Games (EAK, YPH, MFS, VC), pp. 2493–2501.

ICML-2018-KaplanisSC #learning

Continual Reinforcement Learning with Complex Synapses (CK, MS, CC), pp. 2502–2511.

ICML-2018-KarmonZG #locality #named

LaVAN: Localized and Visible Adversarial Noise (DK, DZ, YG), pp. 2512–2520.

ICML-2018-KasaiSM #algorithm #analysis #convergence #probability #recursion

Riemannian Stochastic Recursive Gradient Algorithm with Retraction and Vector Transport and Its Convergence Analysis (HK, HS, BM), pp. 2521–2529.

ICML-2018-KatharopoulosF #learning

Not All Samples Are Created Equal: Deep Learning with Importance Sampling (AK, FF), pp. 2530–2539.

ICML-2018-Katz-SamuelsS #identification

Feasible Arm Identification (JKS, CS), pp. 2540–2548.

ICML-2018-0001ZK #constraints #privacy #scalability #summary

Scalable Deletion-Robust Submodular Maximization: Data Summarization with Privacy and Fairness Constraints (EK0, MZ, AK), pp. 2549–2558.

ICML-2018-KeZSLTBPCP #sequence

Focused Hierarchical RNNs for Conditional Sequence Processing (NRK, KZ, AS, ZL, AT, YB, JP, LC, CJP), pp. 2559–2568.

ICML-2018-KearnsNRW #learning

Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness (MJK, SN, AR0, ZSW), pp. 2569–2577.

ICML-2018-KeivaniS #nearest neighbour #using

Improved nearest neighbor search using auxiliary information and priority functions (OK, KS), pp. 2578–2586.

ICML-2018-KennamerKIS #classification #learning #named

ContextNet: Deep learning for Star Galaxy Classification (NK, DK, ATI, FJSL), pp. 2587–2595.

ICML-2018-KerdreuxPd

Frank-Wolfe with Subsampling Oracle (TK, FP, Ad), pp. 2596–2605.

ICML-2018-KhamaruW #convergence #optimisation #problem

Convergence guarantees for a class of non-convex and non-smooth optimization problems (KK, MJW), pp. 2606–2615.

ICML-2018-KhanNTLGS #learning #performance #scalability

Fast and Scalable Bayesian Deep Learning by Weight-Perturbation in Adam (MEK, DN, VT, WL, YG, AS), pp. 2616–2625.

ICML-2018-KhrulkovO #generative #geometry #network

Geometry Score: A Method For Comparing Generative Adversarial Networks (VK, IVO), pp. 2626–2634.

ICML-2018-KilbertusGKVGW

Blind Justice: Fairness with Encrypted Sensitive Attributes (NK, AG, MJK, MV, KPG, AW), pp. 2635–2644.

ICML-2018-Kim #markov #modelling #process

Markov Modulated Gaussian Cox Processes for Semi-Stationary Intensity Modeling of Events Data (MK), pp. 2645–2653.

ICML-2018-KimM

Disentangling by Factorising (HK, AM), pp. 2654–2663.

ICML-2018-KimW #approximate #bound #predict #self #string

Self-Bounded Prediction Suffix Tree via Approximate String Matching (DK0, CJW), pp. 2664–2672.

ICML-2018-KimWGCWVS #concept #testing

Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV) (BK, MW, JG, CJC, JW, FBV, RS), pp. 2673–2682.

ICML-2018-KimWMSR

Semi-Amortized Variational Autoencoders (YK, SW, ACM, DAS, AMR), pp. 2683–2692.

ICML-2018-KipfFWWZ #relational

Neural Relational Inference for Interacting Systems (TNK, EF, KCW, MW, RSZ), pp. 2693–2702.

ICML-2018-KleinbergLY #question

An Alternative View: When Does SGD Escape Local Minima? (RK, YL, YY), pp. 2703–2712.

ICML-2018-KleindessnerA #crowdsourcing

Crowdsourcing with Arbitrary Adversaries (MK, PA), pp. 2713–2722.

ICML-2018-KnoblauchD #detection #online

Spatio-temporal Bayesian On-line Changepoint Detection with Model Selection (JK, TD), pp. 2723–2732.

ICML-2018-KolarijaniEK #exponential #framework #hybrid #performance

Fast Gradient-Based Methods with Exponential Rate: A Hybrid Control Framework (ASK, PME, TK), pp. 2733–2741.

ICML-2018-KomiyamaTHS #constraints #optimisation

Nonconvex Optimization for Regression with Fairness Constraints (JK, AT, JH, HS), pp. 2742–2751.

ICML-2018-KondorT #network #on the

On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups (RK, ST), pp. 2752–2760.

ICML-2018-Koriche #combinator #compilation #game studies #predict

Compiling Combinatorial Prediction Games (FK), pp. 2761–2770.

ICML-2018-KrauseK0R #evaluation #modelling #sequence

Dynamic Evaluation of Neural Sequence Models (BK, EK, IM0, SR), pp. 2771–2780.

ICML-2018-KrishnamurthyWS

Semiparametric Contextual Bandits (AK, ZSW, VS), pp. 2781–2790.

ICML-2018-KuhnleSCT #integer #performance

Fast Maximization of Non-Submodular, Monotonic Functions on the Integer Lattice (AK, JDS, VGC, MTT), pp. 2791–2800.

ICML-2018-KuleshovFE #learning #nondeterminism #using

Accurate Uncertainties for Deep Learning Using Calibrated Regression (VK, NF, SE), pp. 2801–2809.

ICML-2018-KumarSJ #kernel #metric #network

Trainable Calibration Measures For Neural Networks From Kernel Mean Embeddings (AK, SS, UJ), pp. 2810–2819.

ICML-2018-KuzborskijL #probability

Data-Dependent Stability of Stochastic Gradient Descent (IK, CHL), pp. 2820–2829.

ICML-2018-LiGD #bias #induction #learning #network

Explicit Inductive Bias for Transfer Learning with Convolutional Networks (XL0, YG, FD), pp. 2830–2839.

ICML-2018-LiangSLS #classification #comprehension #network

Understanding the Loss Surface of Neural Networks for Binary Classification (SL, RS, YL, RS), pp. 2840–2849.

ICML-2018-LucasTOV #symmetry

Mixed batches and symmetric discriminators for GAN training (TL, CT, YO, JV), pp. 2850–2859.

ICML-2018-LaberMP #approximate

Binary Partitions with Approximate Minimum Impurity (ESL, MM, FdAMP), pp. 2860–2868.

ICML-2018-LacroixUO #canonical #composition #knowledge base

Canonical Tensor Decomposition for Knowledge Base Completion (TL, NU, GO), pp. 2869–2878.

ICML-2018-LakeB #composition #network

Generalization without Systematicity: On the Compositional Skills of Sequence-to-Sequence Recurrent Networks (BML, MB), pp. 2879–2888.

ICML-2018-LanCS #analysis #estimation #framework

An Estimation and Analysis Framework for the Rasch Model (ASL, MC, CS), pp. 2889–2897.

ICML-2018-LangeKA #bound #clustering #correlation #performance

Partial Optimality and Fast Lower Bounds for Weighted Correlation Clustering (JHL, AK, BA), pp. 2898–2907.

ICML-2018-LaurentB #linear #network

Deep Linear Networks with Arbitrary Loss: All Local Minima Are Global (TL0, JvB), pp. 2908–2913.

ICML-2018-LaurentB18a #multi #network

The Multilinear Structure of ReLU Networks (TL0, JvB), pp. 2914–2922.

ICML-2018-0001JADYD #learning

Hierarchical Imitation and Reinforcement Learning (HML0, NJ, AA, MD, YY, HDI), pp. 2923–2932.

ICML-2018-LeeC #metric

Gradient-Based Meta-Learning with Learned Layerwise Metric and Subspace (YL, SC), pp. 2933–2942.

ICML-2018-LeeKCL #case study #game studies #learning

Deep Reinforcement Learning in Continuous Action Spaces: a Case Study in the Game of Simulated Curling (KL, SAK, JC, SWL), pp. 2943–2952.

ICML-2018-LeePCXS #network

Gated Path Planning Networks (LL, EP, DSC, EPX, RS), pp. 2953–2961.

ICML-2018-LeeYH #learning #multi #symmetry

Deep Asymmetric Multi-task Feature Learning (HL, EY, SJH), pp. 2962–2970.

ICML-2018-LehtinenMHLKAA #image #learning #named

Noise2Noise: Learning Image Restoration without Clean Data (JL, JM, JH, SL, TK, MA, TA), pp. 2971–2980.

ICML-2018-LevinRMP #graph

Out-of-sample extension of graph adjacency spectral embedding (KL, FRK, MWM, CEP), pp. 2981–2990.

ICML-2018-LiH #approach #learning #network

An Optimal Control Approach to Deep Learning and Applications to Discrete-Weight Neural Networks (QL, SH), pp. 2991–3000.

ICML-2018-LiHT0QWL #robust #towards

Towards Binary-Valued Gates for Robust LSTM Training (ZL, DH, FT, WC0, TQ, LW0, TYL), pp. 3001–3010.

ICML-2018-0001MPS #approximate #first-order #on the

On the Limitations of First-Order Approximation in GAN Dynamics (JL0, AM, JP, LS), pp. 3011–3019.

ICML-2018-LiM #clustering

Submodular Hypergraphs: p-Laplacians, Cheeger Inequalities and Spectral Clustering (PL0, OM), pp. 3020–3029.

ICML-2018-LiS

The Well-Tempered Lasso (YL, YS), pp. 3030–3038.

ICML-2018-LiWZ #estimation #markov

Estimation of Markov Chain via Rank-constrained Likelihood (XL, MW, AZ), pp. 3039–3048.

ICML-2018-LianZZL #distributed #parallel #probability

Asynchronous Decentralized Parallel Stochastic Gradient Descent (XL, WZ0, CZ, JL0), pp. 3049–3058.

ICML-2018-LiangLNMFGGJS #abstraction #distributed #learning #named

RLlib: Abstractions for Distributed Reinforcement Learning (EL, RL, RN, PM, RF, KG, JG, MIJ, IS), pp. 3059–3068.

ICML-2018-LiaoC #on the #random

On the Spectrum of Random Features Maps of High Dimensional Data (ZL, RC), pp. 3069–3077.

ICML-2018-LiaoC18a #approach #learning #matrix #random

The Dynamics of Learning: A Random Matrix Approach (ZL, RC), pp. 3078–3087.

ICML-2018-LiaoXFZYPUZ

Reviving and Improving Recurrent Back-Propagation (RL, YX, EF, LZ, KY, XP, RU, RSZ), pp. 3088–3097.

ICML-2018-LinC #distributed #learning #multi #probability

Optimal Distributed Learning with Multi-pass Stochastic Gradient Methods (JL, VC), pp. 3098–3107.

ICML-2018-LinC18a #algorithm #sketching

Optimal Rates of Sketched-regularized Algorithms for Least-Squares Regression over Hilbert Spaces (JL, VC), pp. 3108–3117.

ICML-2018-LinMY #optimisation

Level-Set Methods for Finite-Sum Constrained Convex Optimization (QL, RM, TY), pp. 3118–3127.

ICML-2018-LiptonWS #black box #detection #predict

Detecting and Correcting for Label Shift with Black Box Predictors (ZCL, YXW, AJS), pp. 3128–3136.

ICML-2018-LiuCWO #process #robust #scalability

Generalized Robust Bayesian Committee Machine for Large-scale Gaussian Process Regression (HL, JC, YW, YSO), pp. 3137–3146.

ICML-2018-LiuDLLRS #black box #education #towards

Towards Black-box Iterative Machine Teaching (WL, BD, XL, ZL0, JMR, LS), pp. 3147–3155.

ICML-2018-LiuDRSH #machine learning

Delayed Impact of Fair Machine Learning (LTL, SD, ER, MS, MH), pp. 3156–3164.

ICML-2018-LiuGS #distance

A Two-Step Computation of the Exact GAN Wasserstein Distance (HL, XG, DS), pp. 3165–3174.

ICML-2018-LiuGDFH #detection

Open Category Detection with PAC Guarantees (SL, RG, TGD, AF, DH), pp. 3175–3184.

ICML-2018-LiuH #performance #probability #reduction

Fast Variance Reduction Method with Stochastic Batch Size (XL, CJH), pp. 3185–3194.

ICML-2018-LiuZCWY #performance #probability

Fast Stochastic AUC Maximization with O(1/n)-Convergence Rate (ML, XZ, ZC, XW, TY), pp. 3195–3203.

ICML-2018-LocatelloRKRSSJ #coordination #on the

On Matching Pursuit and Coordinate Descent (FL, AR, SPK, GR, BS, SUS, MJ), pp. 3204–3213.

ICML-2018-LongLMD #learning #named

PDE-Net: Learning PDEs from Data (ZL, YL, XM, BD0), pp. 3214–3222.

ICML-2018-LopesWM #algorithm #estimation #fault #random

Error Estimation for Randomized Least-Squares Algorithms via the Bootstrap (MEL, SW, MWM), pp. 3223–3232.

ICML-2018-LorenziF #modelling #probability

Constraining the Dynamics of Deep Probabilistic Models (ML, MF), pp. 3233–3242.

ICML-2018-LoukasV #approximate #graph #scalability

Spectrally Approximating Large Graphs with Smaller Graphs (AL, PV), pp. 3243–3252.

ICML-2018-LuCLLW #combinator #statistics #trade-off

The Edge Density Barrier: Computational-Statistical Tradeoffs in Combinatorial Inference (HL, YC, JL, HL0, ZW), pp. 3253–3262.

ICML-2018-LuFM #coordination

Accelerating Greedy Coordinate Descent Methods (HL, RMF, VSM), pp. 3263–3272.

ICML-2018-LuGDL #optimisation

Structured Variationally Auto-encoded Optimization (XL, JG, ZD, NDL), pp. 3273–3281.

ICML-2018-LuZLD #architecture #difference #equation #finite #network

Beyond Finite Layer Neural Networks: Bridging Deep Architectures and Numerical Differential Equations (YL, AZ, QL, BD0), pp. 3282–3291.

ICML-2018-LuoSZLZW #learning

End-to-end Active Object Tracking via Reinforcement Learning (WL, PS, FZ, WL0, TZ0, YW), pp. 3292–3301.

ICML-2018-LykourisV

Competitive Caching with Machine Learned Advice (TL, SV), pp. 3302–3311.

ICML-2018-Lyu0YZ0 #automation #design #multi #optimisation

Batch Bayesian Optimization via Multi-objective Acquisition Ensemble for Automated Analog Circuit Design (WL, FY0, CY, DZ, XZ0), pp. 3312–3320.

ICML-2018-MassiasSG #named #performance

Celer: a Fast Solver for the Lasso with Dual Extrapolation (MM, JS, AG), pp. 3321–3330.

ICML-2018-MaBB #comprehension #effectiveness #learning #power of

The Power of Interpolation: Understanding the Effectiveness of SGD in Modern Over-parametrized Learning (SM, RB, MB), pp. 3331–3340.

ICML-2018-MaOSS #matrix

Gradient Descent for Sparse Rank-One Matrix Completion for Crowd-Sourced Aggregation of Sparsely Interacting Workers (YM, AO, CS, VS), pp. 3341–3350.

ICML-2018-MaWCC #estimation #matrix #retrieval #statistics

Implicit Regularization in Nonconvex Statistical Estimation: Gradient Descent Converges Linearly for Phase Retrieval and Matrix Completion (CM, KW, YC, YC0), pp. 3351–3360.

ICML-2018-MaWHZEXWB #learning

Dimensionality-Driven Learning with Noisy Labels (XM, YW0, MEH, SZ0, SME, STX, SNRW, JB0), pp. 3361–3370.

ICML-2018-MaXM #approximate #message passing #optimisation

Approximate message passing for amplitude based optimization (JM0, JX, AM), pp. 3371–3380.

ICML-2018-MadrasCPZ #learning

Learning Adversarially Fair and Transferable Representations (DM, EC, TP, RSZ), pp. 3381–3390.

ICML-2018-MalikPFHRD #learning #performance

An Efficient, Generalized Bellman Update For Cooperative Inverse Reinforcement Learning (DM, MP, JFF, DHM, SJR, ADD), pp. 3391–3399.

ICML-2018-MarinoYM

Iterative Amortized Inference (JM, YY, SM), pp. 3400–3409.

ICML-2018-MarinovMA #analysis #component #streaming

Streaming Principal Component Analysis in Noisy Settings (TVM, PM, RA), pp. 3410–3419.

ICML-2018-MartinLV #approximate #clustering #network #performance

Fast Approximate Spectral Clustering for Dynamic Networks (LM, AL, PV), pp. 3420–3429.

ICML-2018-MazharRFH #clustering #detection

Bayesian Model Selection for Change Point Detection and Clustering (OM, CRR, CF, MRH), pp. 3430–3439.

ICML-2018-McLeodRO #optimisation #performance

Optimization, Fast and Slow: Optimally Switching between Local and Bayesian Optimization (MM, SJR, MAO), pp. 3440–3449.

ICML-2018-MehrabiTY #approximate #bound #network #power of

Bounds on the Approximation Power of Feedforward Neural Networks (MM, AT, MIY), pp. 3450–3458.

ICML-2018-MenschB #predict #programming

Differentiable Dynamic Programming for Structured Prediction and Attention (AM, MB), pp. 3459–3468.

ICML-2018-Mesaoudi-PaulHB #ranking #sorting

Ranking Distributions based on Noisy Sorting (AEMP, EH, RBF), pp. 3469–3477.

ICML-2018-MeschederGN #question

Which Training Methods for GANs do actually Converge? (LMM, AG, SN), pp. 3478–3487.

ICML-2018-MetelliMR #configuration management #markov #process

Configurable Markov Decision Processes (AMM, MM, MR), pp. 3488–3497.

ICML-2018-MetzlerSVB #flexibility #named #network #retrieval #robust

prDeep: Robust Phase Retrieval with a Flexible Deep Network (CAM, PS, AV, RGB), pp. 3498–3507.

ICML-2018-MeyersonM #learning #multi #pseudo

Pseudo-task Augmentation: From Deep Multitask Learning to Intratask Sharing-and Back (EM, RM), pp. 3508–3517.

ICML-2018-MhamdiGR #distributed #learning

The Hidden Vulnerability of Distributed Learning in Byzantium (EMEM, RG, SR), pp. 3518–3527.

ICML-2018-MianjyA #probability

Stochastic PCA with 𝓁2 and 𝓁1 Regularization (PM, RA), pp. 3528–3536.

ICML-2018-MianjyAV #bias #on the

On the Implicit Bias of Dropout (PM, RA, RV), pp. 3537–3545.

ICML-2018-MichaelisBE #segmentation

One-Shot Segmentation in Clutter (CM, MB, ASE), pp. 3546–3555.

ICML-2018-MiconiSC #network

Differentiable plasticity: training plastic neural networks with backpropagation (TM, KOS, JC), pp. 3556–3565.

ICML-2018-MirmanDDGV

Training Neural Machines with Trace-Based Supervision (MM, DD, PD, TG, MTV), pp. 3566–3574.

ICML-2018-MirmanGV #abstract interpretation #network #robust

Differentiable Abstract Interpretation for Provably Robust Neural Networks (MM, TG, MTV), pp. 3575–3583.

ICML-2018-MishchenkoIMA #algorithm #distributed #learning

A Delay-tolerant Proximal-Gradient Algorithm for Distributed Learning (KM, FI, JM, MRA), pp. 3584–3592.

ICML-2018-Mitrovic0ZK #approach #scalability #summary

Data Summarization at Scale: A Two-Stage Submodular Approach (MM, EK0, MZ, AK), pp. 3593–3602.

ICML-2018-Moens #adaptation

The Hierarchical Adaptive Forgetting Variational Filter (VM), pp. 3603–3612.

ICML-2018-MokhtariHK #distributed

Decentralized Submodular Maximization: Bridging Discrete and Continuous Settings (AM, HH, AK), pp. 3613–3622.

ICML-2018-MoreauOV #coordination #distributed #named

DICOD: Distributed Convolutional Coordinate Descent for Convolutional Sparse Coding (TM, LO, NV), pp. 3623–3631.

ICML-2018-MorvanV #algorithm #higher-order #interactive #modelling #named #set

WHInter: A Working set algorithm for High-dimensional sparse second order Interaction models (MLM, JPV), pp. 3632–3641.

ICML-2018-MouZGW #bound

Dropout Training, Data-dependent Regularization, and Generalization Bounds (WM, YZ, JG, LW0), pp. 3642–3650.

ICML-2018-MullerMI #kernel #matrix

Kernelized Synaptic Weight Matrices (LKM, JNPM, GI), pp. 3651–3660.

ICML-2018-MunkhdalaiYMT #adaptation #agile

Rapid Adaptation with Conditionally Shifted Neurons (TM, XY, SM, AT), pp. 3661–3670.

ICML-2018-MussmannL #fault #nondeterminism #on the #performance

On the Relationship between Data Efficiency and Error for Uncertainty Sampling (SM, PL), pp. 3671–3679.

ICML-2018-NachmaniPTW

Fitting New Speakers Based on a Short Untranscribed Sample (EN, AP, YT, LW), pp. 3680–3688.

ICML-2018-Nachum0TS #learning #policy

Smoothed Action Value Functions for Learning Gaussian Policies (ON, MN0, GT, DS), pp. 3689–3697.

ICML-2018-NarayanamurthyV #robust

Nearly Optimal Robust Subspace Tracking (PN, NV), pp. 3698–3706.

ICML-2018-NatoleYL #algorithm #probability

Stochastic Proximal Algorithms for AUC Maximization (MN, YY, SL), pp. 3707–3716.

ICML-2018-NeelR #adaptation #bias #difference #privacy

Mitigating Bias in Adaptive Data Gathering via Differential Privacy (SN, AR0), pp. 3717–3726.

ICML-2018-Nguyen0 #optimisation

Optimization Landscape and Expressivity of Deep CNNs (QN0, MH0), pp. 3727–3736.

ICML-2018-NguyenM0 #network

Neural Networks Should Be Wide Enough to Learn Disconnected Decision Regions (QN0, MCM, MH0), pp. 3737–3746.

ICML-2018-NguyenNDRST #bound #convergence #exclamation

SGD and Hogwild! Convergence Without the Bounded Gradients Assumption (LMN, PHN, MvD, PR, KS, MT), pp. 3747–3755.

ICML-2018-NguyenRF #framework #performance #robust #testing

Active Testing: An Efficient and Robust Framework for Estimating Accuracy (PXN, DR, CCF), pp. 3756–3765.

ICML-2018-NguyenSH #learning #on the

On Learning Sparsely Used Dictionaries from Incomplete Samples (TVN, AS, CH), pp. 3766–3775.

ICML-2018-NickelK #geometry #learning

Learning Continuous Hierarchies in the Lorentz Model of Hyperbolic Geometry (MN, DK), pp. 3776–3785.

ICML-2018-NickischSG #process

State Space Gaussian Processes with Non-Gaussian Likelihood (HN, AS, AG), pp. 3786–3795.

ICML-2018-NiculaeMBC #named

SparseMAP: Differentiable Sparse Structured Inference (VN, AFTM, MB, CC), pp. 3796–3805.

ICML-2018-NieZP #behaviour #visualisation

A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations (WN, YZ, AP), pp. 3806–3815.

ICML-2018-NitandaS #functional #network

Functional Gradient Boosting based on Residual Network Perception (AN, TS), pp. 3816–3825.

ICML-2018-Norouzi-FardTMZ #approximate #data type

Beyond 1/2-Approximation for Submodular Maximization on Massive Data Streams (ANF, JT, SM, AZ, AM, OS), pp. 3826–3835.

ICML-2018-ODonoghueOMM #equation #nondeterminism

The Uncertainty Bellman Equation and Exploration (BO, IO, RM, VM), pp. 3836–3845.

ICML-2018-OdenaBOBORG #generative #performance #question

Is Generator Conditioning Causally Related to GAN Performance? (AO, JB, CO, TBB, CO, CR, IJG), pp. 3846–3855.

ICML-2018-OglicG #kernel #learning

Learning in Reproducing Kernel Krein Spaces (DO, TG0), pp. 3856–3864.

ICML-2018-OhGW #kernel #optimisation

BOCK : Bayesian Optimization with Cylindrical Kernels (CO, EG, MW), pp. 3865–3874.

ICML-2018-OhGSL #learning #self

Self-Imitation Learning (JO, YG, SS, HL), pp. 3875–3884.

ICML-2018-OkunoHS #framework #learning #multi #network #probability

A probabilistic framework for multi-view feature learning with many-to-many associations via neural networks (AO, TH, HS), pp. 3885–3894.

ICML-2018-OlivaDZPSXS #network

Transformation Autoregressive Networks (JBO, AD, MZ, BP, RS, EPX, JS), pp. 3895–3904.

ICML-2018-OlofssonDM #data-driven #design

Design of Experiments for Model Discrimination Hybridising Analytical and Data-Driven Approaches (SO, MPD, RM), pp. 3905–3914.

ICML-2018-OordLBSVKDLCSCG #parallel #performance #speech #synthesis

Parallel WaveNet: Fast High-Fidelity Speech Synthesis (AvdO, YL, IB, KS, OV, KK, GvdD, EL, LCC, FS, NC, DG, SN, SD, EE, NK, HZ, AG, HK, TW, DB, DH), pp. 3915–3923.

ICML-2018-OsamaZS #learning #locality #modelling #streaming

Learning Localized Spatio-Temporal Models From Streaming Data (MO, DZ, TBS), pp. 3924–3932.

ICML-2018-OstrovskiDM #generative #modelling #network

Autoregressive Quantile Networks for Generative Modeling (GO, WD, RM), pp. 3933–3942.

ICML-2018-OstrovskiiH #adaptation #algorithm #first-order #performance

Efficient First-Order Algorithms for Adaptive Signal Denoising (DO, ZH), pp. 3943–3952.

ICML-2018-OttAGR #nondeterminism

Analyzing Uncertainty in Neural Machine Translation (MO, MA, DG, MR), pp. 3953–3962.

ICML-2018-Oymak #learning #network

Learning Compact Neural Networks with Regularization (SO), pp. 3963–3972.

ICML-2018-PaassenGMH #adaptation #distance #edit distance #learning

Tree Edit Distance Learning via Adaptive Symbol Embeddings (BP, CG, AM, BH), pp. 3973–3982.

ICML-2018-PanFWNGN #difference #equation #learning

Reinforcement Learning with Function-Valued Action Spaces for Partial Differential Equation Control (YP, AmF, MW, SN, PG, DN), pp. 3983–3992.

ICML-2018-PanS #learning #predict

Learning to Speed Up Structured Output Prediction (XP, VS), pp. 3993–4002.

ICML-2018-PanZD #analysis #learning

Theoretical Analysis of Image-to-Image Translation with Adversarial Learning (XP, MZ, DD), pp. 4003–4012.

ICML-2018-PangDZ #analysis #linear #network

Max-Mahalanobis Linear Discriminant Analysis Networks (TP, CD, JZ0), pp. 4013–4022.

ICML-2018-PapiniBCPR #policy #probability

Stochastic Variance-Reduced Policy Gradient (MP, DB, GC, MP, MR), pp. 4023–4032.

ICML-2018-ParascandoloKRS #independence #learning

Learning Independent Causal Mechanisms (GP, NK, MRC, BS), pp. 4033–4041.

ICML-2018-PardoTLK #learning

Time Limits in Reinforcement Learning (FP, AT, VL, PK), pp. 4042–4051.

ICML-2018-ParmarVUKSKT #image

Image Transformer (NP, AV, JU, LK, NS, AK, DT), pp. 4052–4061.

ICML-2018-ParmasR0D #flexibility #modelling #named #policy #robust

PIPPS: Flexible Model-Based Policy Search Robust to the Curse of Chaos (PP, CER, JP0, KD), pp. 4062–4071.

ICML-2018-PearceBZN #approach #learning #predict

High-Quality Prediction Intervals for Deep Learning: A Distribution-Free, Ensembled Approach (TP, AB, MZ, AN), pp. 4072–4081.

ICML-2018-PedregosaG #adaptation

Adaptive Three Operator Splitting (FP, GG), pp. 4082–4091.

ICML-2018-PhamGZLD #architecture #parametricity #performance

Efficient Neural Architecture Search via Parameter Sharing (HP, MYG, BZ, QVL, JD), pp. 4092–4101.

ICML-2018-Pike-Burke0SG #feedback

Bandits with Delayed, Aggregated Anonymous Feedback (CPB, SA0, CS, SG), pp. 4102–4110.

ICML-2018-PleissGWW #predict #process

Constant-Time Predictive Distributions for Gaussian Processes (GP, JRG, KQW, AGW), pp. 4111–4120.

ICML-2018-PoonLS #convergence

Local Convergence Properties of SAGA/Prox-SVRG and Acceleration (CP, JL, CBS), pp. 4121–4129.

ICML-2018-Pouliot #equivalence #multi #performance #statistics

Equivalence of Multicategory SVM and Simplex Cone SVM: Fast Computations and Statistical Theory (GP), pp. 4130–4137.

ICML-2018-PretoriusKK #learning #linear

Learning Dynamics of Linear Denoising Autoencoders (AP, SK, HK), pp. 4138–4147.

ICML-2018-PuDGWWZHC #generative #learning #multi #named

JointGAN: Multi-Domain Joint Distribution Learning with Generative Adversarial Nets (YP, SD, ZG, WW, GW0, YZ, RH, LC), pp. 4148–4157.

ICML-2018-PuMSK #synthesis

Selecting Representative Examples for Program Synthesis (YP, ZM, ASL, LPK), pp. 4158–4167.

ICML-2018-QiJZ #earley #parsing #predict #sequence

Generalized Earley Parser: Bridging Symbolic Grammars and Sequence Data for Future Prediction (SQ, BJ, SCZ), pp. 4168–4176.

ICML-2018-Qiao #collaboration #question

Do Outliers Ruin Collaboration? (MQ), pp. 4177–4184.

ICML-2018-QiaoZ0WY #image #network #recognition #scalability

Gradually Updated Neural Networks for Large-Scale Image Recognition (SQ, ZZ, WS0, BW0, ALY), pp. 4185–4194.

ICML-2018-QiuCCS #named #network

DCFNet: Deep Neural Network with Decomposed Convolutional Filters (QQ, XC, ARC, GS), pp. 4195–4204.

ICML-2018-QuLX

Non-convex Conditional Gradient Sliding (CQ, YL, HX), pp. 4205–4214.

ICML-2018-RabinowitzPSZEB

Machine Theory of Mind (NCR, FP, HFS, CZ, SMAE, MB), pp. 4215–4224.

ICML-2018-RaeDDL #learning #parametricity #performance

Fast Parametric Learning with Activation Memorization (JWR, CD, PD, TPL), pp. 4225–4234.

ICML-2018-RaghuIAKLK #game studies #learning #question

Can Deep Reinforcement Learning Solve Erdos-Selfridge-Spencer Games? (MR, AI, JA, RK, QVL, JMK), pp. 4235–4243.

ICML-2018-RaguetL #algorithm #graph

Cut-Pursuit Algorithm for Regularizing Nonsmooth Functionals with Graph Total Variation (HR, LL), pp. 4244–4253.

ICML-2018-RaileanuDSF #learning #modelling #multi #using

Modeling Others using Oneself in Multi-Agent Reinforcement Learning (RR, ED, AS, RF), pp. 4254–4263.

ICML-2018-RainforthCYW #monte carlo #on the

On Nesting Monte Carlo Estimators (TR, RC, HY, AW), pp. 4264–4273.

ICML-2018-RainforthKLMIWT #bound

Tighter Variational Bounds are Not Necessarily Better (TR, ARK, TAL, CJM, MI, FW, YWT), pp. 4274–4282.

ICML-2018-RamdasZWJ #adaptation #algorithm #named #online

SAFFRON: an Adaptive Algorithm for Online Control of the False Discovery Rate (AR, TZ, MJW, MIJ), pp. 4283–4291.

ICML-2018-RashidSWFFW #learning #multi #named

QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning (TR, MS, CSdW, GF, JNF, SW), pp. 4292–4301.

ICML-2018-RavivTDT #graph

Gradient Coding from Cyclic MDS Codes and Expander Graphs (NR, RT, AD, IT), pp. 4302–4310.

ICML-2018-RavuriMRV #generative #learning #modelling

Learning Implicit Generative Models with the Method of Learned Moments (SVR, SM, MR, OV), pp. 4311–4320.

ICML-2018-ReagenGAMRWB #encoding #named #network

Weightless: Lossy weight encoding for deep neural network compression (BR, UG, BA, MM, AMR, GYW, DB0), pp. 4321–4330.

ICML-2018-RenZYU #learning #robust

Learning to Reweight Examples for Robust Deep Learning (MR, WZ, BY, RU), pp. 4331–4340.

ICML-2018-RiedmillerHLNDW #game studies #learning

Learning by Playing Solving Sparse Reward Tasks from Scratch (MAR, RH, TL, MN, JD, TVdW, VM, NH, JTS), pp. 4341–4350.

ICML-2018-RitterWKJBPB

Been There, Done That: Meta-Learning with Episodic Recall (SR, JXW, ZKN, SMJ, CB, RP, MB), pp. 4351–4360.

ICML-2018-RobertsERHE #learning #music

A Hierarchical Latent Vector Model for Learning Long-Term Structure in Music (AR, JHE, CR, CH, DE), pp. 4361–4370.

ICML-2018-RosenfeldBGS #combinator #learning

Learning to Optimize Combinatorial Functions (NR, EB, AG, YS), pp. 4371–4380.

ICML-2018-RuMGO #optimisation #performance

Fast Information-theoretic Bayesian Optimisation (BXR, MM, DG, MAO), pp. 4381–4389.

ICML-2018-RuffGDSVBMK #classification

Deep One-Class Classification (LR, NG, LD, SAS, RAV, AB, EM, MK), pp. 4390–4399.

ICML-2018-RuizTDB #category theory #probability #scalability

Augment and Reduce: Stochastic Inference for Large Categorical Distributions (FJRR, MKT, ABD, DMB), pp. 4400–4409.

ICML-2018-RukatHY #composition #probability

Probabilistic Boolean Tensor Decomposition (TR, CCH, CY), pp. 4410–4419.

ICML-2018-RyderGMP #black box #difference #equation #probability

Black-Box Variational Inference for Stochastic Differential Equations (TR, AG, ASM, DP), pp. 4420–4429.

ICML-2018-SafranS #network

Spurious Local Minima are Common in Two-Layer ReLU Neural Networks (IS, OS), pp. 4430–4438.

ICML-2018-SahooLM #equation #learning

Learning Equations for Extrapolation and Control (SSS, CHL, GM), pp. 4439–4447.

ICML-2018-SajjadiPMS #network

Tempered Adversarial Networks (MSMS, GP, AM, BS), pp. 4448–4456.

ICML-2018-SalaSGR #representation #trade-off

Representation Tradeoffs for Hyperbolic Embeddings (FS, CDS, AG, CR), pp. 4457–4466.

ICML-2018-Sanchez-Gonzalez #graph #network #physics

Graph Networks as Learnable Physics Engines for Inference and Control (ASG, NH, JTS, JM, MAR, RH, PWB), pp. 4467–4476.

ICML-2018-SantoroHBML #network #reasoning

Measuring abstract reasoning in neural networks (AS, FH, DGTB, ASM, TPL), pp. 4477–4486.

ICML-2018-SanturkarSM #classification

A Classification-Based Study of Covariate Shift in GAN Distributions (SS, LS, AM), pp. 4487–4496.

ICML-2018-SanyalKGK #as a service #named #predict

TAPAS: Tricks to Accelerate (encrypted) Prediction As a Service (AS, MJK, AG, VK), pp. 4497–4506.

ICML-2018-Scarlett #bound #optimisation

Tight Regret Bounds for Bayesian Optimization in One Dimension (JS), pp. 4507–4515.

ICML-2018-SchmitJ #learning

Learning with Abandonment (SS, RJ), pp. 4516–4524.

ICML-2018-SchwabKMMSK #learning #multi

Not to Cry Wolf: Distantly Supervised Multitask Learning in Critical Care (PS, EK, CM, DJM, CS, WK), pp. 4525–4534.

ICML-2018-Schwarz0LGTPH #framework #learning #scalability

Progress & Compress: A scalable framework for continual learning (JS, WC0, JL, AGB, YWT, RP, RH), pp. 4535–4544.

ICML-2018-SenKS #black box #multi #optimisation

Multi-Fidelity Black-Box Optimization with Hierarchical Partitions (RS, KK, SS), pp. 4545–4554.

ICML-2018-SerraSMK

Overcoming Catastrophic Forgetting with Hard Attention to the Task (JS, DS, MM, AK), pp. 4555–4564.

ICML-2018-SerraTR #bound #linear #network

Bounding and Counting Linear Regions of Deep Neural Networks (TS, CT, SR), pp. 4565–4573.

ICML-2018-SewardUBJH #first-order #generative #network

First Order Generative Adversarial Networks (CS, TU, UB, NJ, SH), pp. 4574–4583.

ICML-2018-SharchilevUSR

Finding Influential Training Samples for Gradient Boosted Decision Trees (BS, YU, PS, MdR), pp. 4584–4592.

ICML-2018-SharmaNK #clique #problem #random #using

Solving Partial Assignment Problems using Random Clique Complexes (CS, DN, MK), pp. 4593–4602.

ICML-2018-ShazeerS #adaptation #learning #memory management #named #sublinear

Adafactor: Adaptive Learning Rates with Sublinear Memory Cost (NS, MS), pp. 4603–4611.

ICML-2018-Sheffet #testing

Locally Private Hypothesis Testing (OS), pp. 4612–4621.

ICML-2018-SheldonWS #automation #difference #integer #learning #modelling

Learning in Integer Latent Variable Models with Nested Automatic Differentiation (DS, KW, DS), pp. 4622–4630.

ICML-2018-ShenMZZQ #communication #convergence #distributed #learning #performance #probability #towards

Towards More Efficient Stochastic Decentralized Learning: Faster Convergence and Sparse Communication (ZS, AM, TZ, PZ, HQ), pp. 4631–4640.

ICML-2018-ShenSWLZ #algorithm #framework #hybrid #metric

An Algorithmic Framework of Variable Metric Over-Relaxed Hybrid Proximal Extra-Gradient Method (LS, PS, YW, WL0, TZ0), pp. 4641–4650.

ICML-2018-ShiS0 #approach #estimation

A Spectral Approach to Gradient Estimation for Implicit Distributions (JS, SS, JZ0), pp. 4651–4660.

ICML-2018-ShiarlisWSWP #composition #learning #named

TACO: Learning Task Decomposition via Temporal Alignment for Control (KS, MW, SS, SW, IP), pp. 4661–4670.

ICML-2018-SibliniMK #clustering #learning #multi #performance #random

CRAFTML, an Efficient Clustering-based Random Forest for Extreme Multi-label Learning (WS, FM, PK), pp. 4671–4680.

ICML-2018-SimsekliYNCR #optimisation #probability

Asynchronous Stochastic Quasi-Newton MCMC for Non-Convex Optimization (US, CY, THN, ATC, GR), pp. 4681–4690.

ICML-2018-Sinha #clustering #matrix #random #using

K-means clustering using random matrix sparsification (KS), pp. 4691–4699.

ICML-2018-Skerry-RyanBXWS #speech #synthesis #towards

Towards End-to-End Prosody Transfer for Expressive Speech Synthesis with Tacotron (RJSR, EB, YX, YW, DS, JS, RJW, RC, RAS), pp. 4700–4709.

ICML-2018-SmithHP #learning #policy

An Inference-Based Policy Gradient Method for Learning Options (MS, HvH, JP), pp. 4710–4719.

ICML-2018-SongSE #higher-order

Accelerating Natural Gradient with Higher-Order Invariance (YS, JS, SE), pp. 4720–4729.

ICML-2018-SrinivasF #information management

Knowledge Transfer with Jacobian Matching (SS, FF), pp. 4730–4738.

ICML-2018-SrinivasJALF #learning #network

Universal Planning Networks: Learning Generalizable Representations for Visuomotor Control (AS, AJ, PA, SL, CF), pp. 4739–4748.

ICML-2018-SroujiZS #learning

Structured Control Nets for Deep Reinforcement Learning (MS, JZ, RS), pp. 4749–4758.

ICML-2018-Streeter #algorithm #approximate #modelling #predict

Approximation Algorithms for Cascading Prediction Models (MS), pp. 4759–4767.

ICML-2018-SuW #learning

Learning Low-Dimensional Temporal Representations (BS, YW), pp. 4768–4777.

ICML-2018-SuganumaOO #image #search-based #standard

Exploiting the Potential of Standard Convolutional Autoencoders for Image Restoration by Evolutionary Search (MS, MO, TO), pp. 4778–4787.

ICML-2018-SuiZBY #optimisation #process

Stagewise Safe Bayesian Optimization with Gaussian Processes (YS, VZ, JWB, YY), pp. 4788–4796.

ICML-2018-SunNSL #synthesis

Neural Program Synthesis from Diverse Demonstration Videos (SHS, HN, SS, JJL), pp. 4797–4806.

ICML-2018-SunP #approximate #scalability

Scalable Approximate Bayesian Inference for Particle Tracking Data (RS, LP), pp. 4807–4816.

ICML-2018-SunTLZ #adaptation #optimisation #visual notation

Graphical Nonconvex Optimization via an Adaptive Convex Relaxation (QS, KMT, HL0, TZ0), pp. 4817–4824.

ICML-2018-SunYDB #matrix #network

Convolutional Imputation of Matrix Networks (QS, MY, DLD, SPB), pp. 4825–4834.

ICML-2018-SunZWZLG #composition #kernel #learning #process

Differentiable Compositional Kernel Learning for Gaussian Processes (SS, GZ, CW, WZ, JL, RBG), pp. 4835–4844.

ICML-2018-Talvitie #learning

Learning the Reward Function for a Misspecified Model (ET), pp. 4845–4854.

ICML-2018-TangLYZL #distributed #named

D2: Decentralized Training over Decentralized Data (HT, XL, MY0, CZ, JL0), pp. 4855–4863.

ICML-2018-TaniaiM

Neural Inverse Rendering for General Reflectance Photometric Stereo (TT, TM), pp. 4864–4873.

ICML-2018-TanseyWBR #black box

Black Box FDR (WT, YW, DMB, RR), pp. 4874–4883.

ICML-2018-TaoBZ #dependence #identification #linear

Best Arm Identification in Linear Bandits with Linear Dimension Dependency (CT, SAB, YZ0), pp. 4884–4893.

ICML-2018-TaoCHFC #generative #network

Chi-square Generative Adversarial Network (CT, LC, RH, JF, LC), pp. 4894–4903.

ICML-2018-TaylorSL #automation #convergence #first-order

Lyapunov Functions for First-Order Methods: Tight Automated Convergence Guarantees (AT, BVS, LL), pp. 4904–4913.

ICML-2018-TeyeAS #estimation #network #nondeterminism #normalisation

Bayesian Uncertainty Estimation for Batch Normalized Deep Networks (MT, HA, KS0), pp. 4914–4923.

ICML-2018-ThomasDB #learning

Decoupling Gradient-Like Learning Rules from Representations (PST, CD, EB), pp. 4924–4932.

ICML-2018-TianZZ #learning #named

CoVeR: Learning Covariate-Specific Vector Representations with Tensor Decompositions (KT, TZ, JZ), pp. 4933–4942.

ICML-2018-TirinzoniSPR #learning

Importance Weighted Transfer of Samples in Reinforcement Learning (AT, AS, MP, MR), pp. 4943–4952.

ICML-2018-TongYAV #multi

Adversarial Regression with Multiple Learners (LT, SY, SA, YV), pp. 4953–4961.

ICML-2018-TouatiBPV #approximate #convergence

Convergent TREE BACKUP and RETRACE with Function Approximation (AT, PLB, DP, PV), pp. 4962–4971.

ICML-2018-TrinhDLL #dependence #learning

Learning Longer-term Dependencies in RNNs with Auxiliary Losses (THT, AMD, TL, QVL), pp. 4972–4981.

ICML-2018-TsakirisV #analysis #clustering

Theoretical Analysis of Sparse Subspace Clustering with Missing Entries (MCT, RV), pp. 4982–4991.

ICML-2018-TschannenKA #learning #multi #named

StrassenNets: Deep Learning with a Multiplication Budget (MT, AK, AA), pp. 4992–5001.

ICML-2018-TsuchidaRG

Invariance of Weight Distributions in Rectified MLPs (RT, FRK, MG), pp. 5002–5011.

ICML-2018-TuR #difference #learning #linear #polynomial

Least-Squares Temporal Difference Learning for the Linear Quadratic Regulator (ST, BR), pp. 5012–5021.

ICML-2018-TuckerBGTGL #learning

The Mirage of Action-Dependent Baselines in Reinforcement Learning (GT, SB, SG, RET, ZG, SL), pp. 5022–5031.

ICML-2018-UesatoOKO

Adversarial Risk and the Dangers of Evaluating Against Weak Attacks (JU, BO, PK, AvdO), pp. 5032–5041.

ICML-2018-VahdatMBKA

DVAE++: Discrete Variational Autoencoders with Overlapping Transformations (AV, WGM, ZB, AK, EA), pp. 5042–5051.

ICML-2018-VermaMSKC #learning

Programmatically Interpretable Reinforcement Learning (AV, VM, RS, PK, SC), pp. 5052–5061.

ICML-2018-VogelBC #learning #optimisation #probability #similarity

A Probabilistic Theory of Supervised Similarity Learning for Pointwise ROC Curve Optimization (RV, AB, SC), pp. 5062–5071.

ICML-2018-WeiZHY #learning

Transfer Learning via Learning to Transfer (YW, YZ, JH, QY), pp. 5072–5081.

ICML-2018-WagnerGKM #data type #learning

Semi-Supervised Learning on Data Streams via Temporal Label Propagation (TW, SG, SPK, NM), pp. 5082–5091.

ICML-2018-WalderK #programming #similarity

Neural Dynamic Programming for Musical Self Similarity (CJW, DK0), pp. 5092–5100.

ICML-2018-WangC #combinator

Thompson Sampling for Combinatorial Semi-Bandits (SW, WC), pp. 5101–5109.

ICML-2018-WangGLWY #learning #predict #towards

PredRNN++: Towards A Resolution of the Deep-in-Time Dilemma in Spatiotemporal Predictive Learning (YW, ZG, ML, JW0, PSY), pp. 5110–5119.

ICML-2018-WangJC #nearest neighbour #robust

Analyzing the Robustness of Nearest Neighbors to Adversarial Examples (YW, SJ, KC), pp. 5120–5129.

ICML-2018-WangK #learning #multi

Competitive Multi-agent Inverse Reinforcement Learning with Sub-optimal Demonstrations (XW, DK), pp. 5130–5138.

ICML-2018-WangLS #matrix #multi

Coded Sparse Matrix Multiplication (SW, JL, NBS), pp. 5139–5147.

ICML-2018-WangSQ #learning #modelling #multi #performance #scalability #visual notation

A Fast and Scalable Joint Estimator for Integrating Additional Knowledge in Learning Multiple Related Sparse Gaussian Graphical Models (BW, AS, YQ), pp. 5148–5157.

ICML-2018-WangSL #streaming

Provable Variable Selection for Streaming Features (JW0, JS0, PL0), pp. 5158–5166.

ICML-2018-WangSZRBSXJRS #modelling #speech #synthesis

Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis (YW, DS, YZ, RJSR, EB, JS, YX, YJ, FR, RAS), pp. 5167–5176.

ICML-2018-WangVLGGZ #network

Adversarial Distillation of Bayesian Neural Network Posteriors (KCW, PV, JL, LG, RBG, RSZ), pp. 5177–5186.

ICML-2018-WangWY #multi

Minimax Concave Penalized Multi-Armed Bandit Model with High-Dimensional Convariates (XW, MMW, TY), pp. 5187–5195.

ICML-2018-WangYKN #online #taxonomy

Online Convolutional Sparse Coding with Sample-Dependent Dictionary (YW, QY, JTYK, LMN), pp. 5196–5205.

ICML-2018-WangZ0 #message passing #modelling #visual notation

Stein Variational Message Passing for Continuous Graphical Models (DW, ZZ, QL0), pp. 5206–5214.

ICML-2018-WangZLMM #approximate #parametricity #performance

Approximate Leave-One-Out for Fast Parameter Tuning in High Dimensions (SW, WZ, HL, AM, VSM), pp. 5215–5224.

ICML-2018-WehrmannCB #classification #multi #network

Hierarchical Multi-Label Classification Networks (JW, RC, RCB), pp. 5225–5234.

ICML-2018-WeinshallCA #education #learning #network

Curriculum Learning by Transfer Learning: Theory and Experiments with Deep Networks (DW, GC, DA), pp. 5235–5243.

ICML-2018-WeissGY #automaton #network #query #using

Extracting Automata from Recurrent Neural Networks Using Queries and Counterexamples (GW, YG, EY), pp. 5244–5253.

ICML-2018-WeiszGS #algorithm #approximate #bound #named

LEAPSANDBOUNDS: A Method for Approximately Optimal Algorithm Configuration (GW, AG, CS), pp. 5254–5262.

ICML-2018-WenHSZCL #network #predict #recognition

Deep Predictive Coding Network for Object Recognition (HW, KH, JS, YZ, EC, ZL), pp. 5263–5272.

ICML-2018-WengZCSHDBD #network #performance #robust #towards

Towards Fast Computation of Certified Robustness for ReLU Networks (TWW, HZ0, HC, ZS, CJH, LD, DSB, ISD), pp. 5273–5282.

ICML-2018-WongK

Provable Defenses against Adversarial Examples via the Convex Outer Adversarial Polytope (EW, JZK), pp. 5283–5292.

ICML-2018-WuCN #estimation

Local Density Estimation in High Dimensions (XW, MC, VN), pp. 5293–5301.

ICML-2018-WuGL #adaptation #trade-off

Adaptive Exploration-Exploitation Tradeoff for Opportunistic Bandits (HW, XG, XL0), pp. 5302–5310.

ICML-2018-WuHS #approach #collaboration #named #ranking

SQL-Rank: A Listwise Approach to Collaborative Ranking (LW, CJH, JS), pp. 5311–5320.

ICML-2018-Wu0H0 #distributed #fault #optimisation #scalability

Error Compensated Quantized SGD and its Applications to Large-scale Distributed Optimization (JW, WH0, JH, TZ0), pp. 5321–5329.

ICML-2018-0001JCCJ #robust #using

Reinforcing Adversarial Robustness using Model Confidence Induced by Adversarial Training (XW0, UJ, JC, LC, SJ), pp. 5330–5338.

ICML-2018-WuSHDR #algorithm #probability #programming #semantics

Discrete-Continuous Mixtures in Probabilistic Programming: Generalized Semantics and Inference Algorithms (YW, SS, NH, SD, SJR), pp. 5339–5348.

ICML-2018-WuW

Variance Regularized Counterfactual Risk Minimization via Variational Divergence Minimization (HW, MDW), pp. 5349–5358.

ICML-2018-WuWWWVL #clustering #parametricity

Deep k-Means: Re-Training and Parameter Sharing with Harder Cluster Assignments for Compressing Deep Convolutions (JW, YW, ZW, ZW, AV, YL), pp. 5359–5368.

ICML-2018-XiBG #multi

Bayesian Quadrature for Multiple Related Integrals (XX, FXB, MAG), pp. 5369–5378.

ICML-2018-XiaTTQYL #learning

Model-Level Dual Learning (YX, XT, FT, TQ, NY, TYL), pp. 5379–5388.

ICML-2018-XiaoBSSP #how #network

Dynamical Isometry and a Mean Field Theory of CNNs: How to Train 10, 000-Layer Vanilla Convolutional Neural Networks (LX, YB, JSD, SSS, JP), pp. 5389–5398.

ICML-2018-XieWZX #analysis #distance #learning #metric

Orthogonality-Promoting Distance Metric Learning: Convex Relaxation and Theoretical Analysis (PX, WW, YZ, EPX), pp. 5399–5408.

ICML-2018-XieZZX

Nonoverlap-Promoting Variable Selection (PX, HZ, YZ, EPX), pp. 5409–5418.

ICML-2018-XieZCC #adaptation #learning #semantics

Learning Semantic Representations for Unsupervised Domain Adaptation (SX, ZZ, LC0, CC), pp. 5419–5428.

ICML-2018-Xu #convergence #estimation

Rates of Convergence of Spectral Methods for Graphon Estimation (JX), pp. 5429–5438.

ICML-2018-XuCZ #learning #process

Learning Registered Point Processes from Idiosyncratic Observations (HX, LC, HZ), pp. 5439–5448.

ICML-2018-XuLTSKJ #graph #learning #network #representation

Representation Learning on Graphs with Jumping Knowledge Networks (KX, CL, YT, TS, KiK, SJ), pp. 5449–5458.

ICML-2018-XuLZP #learning

Learning to Explore via Meta-Policy Gradient (TX, QL0, LZ, JP0), pp. 5459–5468.

ICML-2018-XuMBSD #parametricity

Nonparametric Regression with Comparisons: Escaping the Curse of Dimensionality with Ordinal Information (YX, HM, SB, AS, AD), pp. 5469–5478.

ICML-2018-XuSC #divide and conquer #kernel

Optimal Tuning for Divide-and-conquer Kernel Ridge Regression with Massive Data (GX, ZS, GC), pp. 5479–5487.

ICML-2018-XuWG #probability

Continuous and Discrete-time Accelerated Stochastic Mirror Descent for Strongly Convex Functions (PX0, TW0, QG), pp. 5488–5497.

ICML-2018-XuZFLB #learning #semantics

A Semantic Loss Function for Deep Learning with Symbolic Knowledge (JX, ZZ, TF, YL, GVdB), pp. 5498–5507.

ICML-2018-YabeHSIKFK

Causal Bandits with Propagating Inference (AY, DH, HS, SI, NK, TF, KiK), pp. 5508–5516.

ICML-2018-YanCJ #learning

Active Learning with Logged Data (SY, KC, TJ), pp. 5517–5526.

ICML-2018-YanKZR #classification #metric

Binary Classification with Karmic, Threshold-Quasi-Concave Metrics (BY, OK, KZ, PR), pp. 5527–5536.

ICML-2018-YangKU #equivalence #graph #learning

Characterizing and Learning Equivalence Classes of Causal DAGs under Interventions (KDY, AK, CU), pp. 5537–5546.

ICML-2018-YangK #modelling #network #process #relational

Dependent Relational Gamma Process Models for Longitudinal Networks (SY, HK), pp. 5547–5556.

ICML-2018-YangLRN #testing

Goodness-of-fit Testing for Discrete Distributions via Stein Discrepancy (JY, QL, VAR, JN), pp. 5557–5566.

ICML-2018-YangLLZZW #learning #multi

Mean Field Multi-Agent Reinforcement Learning (YY, RL, ML, MZ, WZ0, JW0), pp. 5567–5576.

ICML-2018-YaoVSG

Yes, but Did It Work?: Evaluating Variational Inference (YY, AV, DS, AG), pp. 5577–5586.

ICML-2018-YaratsL #generative

Hierarchical Text Generation and Planning for Strategic Dialogue (DY, ML), pp. 5587–5595.

ICML-2018-YaroslavtsevV #algorithm #clustering #parallel

Massively Parallel Algorithms and Hardness for Single-Linkage Clustering under 𝓁p Distances (GY, AV), pp. 5596–5605.

ICML-2018-YeA #performance

Communication-Computation Efficient Gradient Coding (MY, EA), pp. 5606–5615.

ICML-2018-YeS #approach #network

Variable Selection via Penalized Neural Network: a Drop-Out-One Loss Approach (MY, YS), pp. 5616–5625.

ICML-2018-YenKYHKR #composition #learning #performance #scalability

Loss Decomposition for Fast Learning in Large Output Spaces (IEHY, SK, FXY, DNHR, SK, PR), pp. 5626–5635.

ICML-2018-YinCRB #distributed #learning #statistics #towards

Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates (DY, YC0, KR, PLB), pp. 5636–5645.

ICML-2018-YinZ

Semi-Implicit Variational Inference (MY, MZ), pp. 5646–5655.

ICML-2018-LiM18a

Disentangled Sequential Autoencoder (YL, SM), pp. 5656–5665.

ICML-2018-YonaR #approximate #learning

Probably Approximately Metric-Fair Learning (GY, GNR), pp. 5666–5674.

ICML-2018-YoonJS #generative #named #using

GAIN: Missing Data Imputation using Generative Adversarial Nets (JY, JJ, MvdS), pp. 5675–5684.

ICML-2018-YoonJS18a #dataset #generative #modelling #multi #named #network #predict #using

RadialGAN: Leveraging multiple datasets to improve target-specific predictive models using Generative Adversarial Networks (JY, JJ, MvdS), pp. 5685–5693.

ICML-2018-YouYRHL #generative #graph #modelling #named

GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models (JY, RY, XR, WLH, JL), pp. 5694–5703.

ICML-2018-YuanST #algorithm #clustering #performance

An Efficient Semismooth Newton Based Algorithm for Convex Clustering (YY, DS, KCT), pp. 5704–5712.

ICML-2018-YurtseverFLC #framework #programming

A Conditional Gradient Framework for Composite Convex Minimization with Applications to Semidefinite Programming (AY, OF, FL, VC), pp. 5713–5722.

ICML-2018-ZadikMS #machine learning #orthogonal

Orthogonal Machine Learning: Power and Limitations (IZ, LWM, VS), pp. 5723–5731.

ICML-2018-ZanetteB #bound #identification #learning #problem

Problem Dependent Reinforcement Learning Bounds Which Can Identify Bandit Structure in MDPs (AZ, EB), pp. 5732–5740.

ICML-2018-ZhangCLC #optimisation #policy

Policy Optimization as Wasserstein Gradient Flows (RZ, CC, CL, LC), pp. 5741–5750.

ICML-2018-ZhangDG #induction #matrix #multi #performance

Fast and Sample Efficient Inductive Matrix Completion via Multi-Phase Procrustes Flow (XZ, SSD, QG), pp. 5751–5760.

ICML-2018-ZhangFS #estimation #matrix #scalability

Large-Scale Sparse Inverse Covariance Estimation via Thresholding and Max-Det Matrix Completion (RYZ, SF, SS), pp. 5761–5770.

ICML-2018-ZhangFL #performance

High Performance Zero-Memory Overhead Direct Convolutions (JZ, FF, TML), pp. 5771–5780.

ICML-2018-ZhangHMLZ

Safe Element Screening for Submodular Function Minimization (WZ, BH, LM0, WL0, TZ0), pp. 5781–5790.

ICML-2018-ZhangKL #algorithm #distributed #privacy

Improving the Privacy and Accuracy of ADMM-Based Distributed Algorithms (XZ, MMK, ML), pp. 5791–5800.

ICML-2018-ZhangLD #network #performance

Stabilizing Gradients for Deep Neural Networks via Efficient SVD Parameterization (JZ, QL, ISD), pp. 5801–5809.

ICML-2018-ZhangLSD #dependence #fourier #learning

Learning Long Term Dependencies via Fourier Recurrent Units (JZ, YL, ZS, ISD), pp. 5810–5818.

ICML-2018-ZhangNL #geometry #network

Tropical Geometry of Deep Neural Networks (LZ, GN, LHL), pp. 5819–5827.

ICML-2018-ZhangP #parametricity

Deep Bayesian Nonparametric Tracking (AZ, JWP), pp. 5828–5836.

ICML-2018-ZhangSLSF #composition

Composable Planning with Attributes (AZ, SS, AL, AS, RF), pp. 5837–5846.

ICML-2018-ZhangSDG

Noisy Natural Gradient as Variational Inference (GZ, SS, DD, RBG), pp. 5847–5856.

ICML-2018-ZhangWYG #analysis #matrix #rank

A Primal-Dual Analysis of Global Optimality in Nonconvex Low-Rank Matrix Recovery (XZ, LW, YY, QG), pp. 5857–5866.

ICML-2018-ZhangYL0B #distributed #learning #multi

Fully Decentralized Multi-Agent Reinforcement Learning with Networked Agents (KZ, ZY, HL0, TZ0, TB), pp. 5867–5876.

ICML-2018-ZhangYJZ #adaptation

Dynamic Regret of Strongly Adaptive Methods (LZ0, TY, RJ, ZHZ), pp. 5877–5886.

ICML-2018-ZhaoDBZ #learning #topic #word

Inter and Intra Topic Structure Learning with Word Embeddings (HZ, LD, WLB, MZ), pp. 5887–5896.

ICML-2018-ZhaoKZRL

Adversarially Regularized Autoencoders (JJZ, YK, KZ, AMR, YL), pp. 5897–5906.

ICML-2018-Zhao0FYW #estimation #feature model #learning

MSplit LBI: Realizing Feature Selection and Dense Estimation Simultaneously in Few-shot and Zero-shot Learning (BZ, XS0, YF, YY0, YW), pp. 5907–5916.

ICML-2018-ZhaoX #modelling #random

Composite Marginal Likelihood Methods for Random Utility Models (ZZ, LX), pp. 5917–5926.

ICML-2018-0004K #finite #infinity #lightweight #optimisation #probability

Lightweight Stochastic Optimization for Minimizing Finite Sums with Infinite Data (SZ0, JTYK), pp. 5927–5935.

ICML-2018-ZhengPF #approach #robust

A Robust Approach to Sequential Information Theoretic Planning (SZ, JP, JWFI), pp. 5936–5944.

ICML-2018-ZhitnikovMM #behaviour #statistics

Revealing Common Statistical Behaviors in Heterogeneous Populations (AZ, RM, TM), pp. 5945–5954.

ICML-2018-ZhouF #comprehension #optimisation #performance

Understanding Generalization and Optimization Performance of Deep CNNs (PZ, JF), pp. 5955–5964.

ICML-2018-ZhouMBGYLF #bound #distributed #how #optimisation #question

Distributed Asynchronous Optimization with Unbounded Delays: How Slow Can You Go? (ZZ, PM, NB, PWG, YY, LJL, LFF0), pp. 5965–5974.

ICML-2018-ZhouSC #algorithm #convergence #performance #probability

A Simple Stochastic Variance Reduced Algorithm with Fast Convergence Rates (KZ, FS, JC), pp. 5975–5984.

ICML-2018-ZhouXG #polynomial #probability

Stochastic Variance-Reduced Cubic Regularized Newton Method (DZ, PX0, QG), pp. 5985–5994.

ICML-2018-ZhouZZ #algorithm #performance

Racing Thompson: an Efficient Algorithm for Thompson Sampling with Non-conjugate Priors (YZ, JZ0, JZ), pp. 5995–6003.

ICML-2018-ZhuL #communication #constraints #distributed #parametricity

Distributed Nonparametric Regression under Communication Constraints (YZ, JL), pp. 6004–6012.

ICML-2018-ZhuoLSZCZ #message passing

Message Passing Stein Variational Gradient Descent (JZ, CL0, JS, JZ0, NC, BZ0), pp. 6013–6022.

ICML-2018-ZouXG #monte carlo #probability

Stochastic Variance-Reduced Hamilton Monte Carlo Methods (DZ, PX0, QG), pp. 6023–6032.

ICML-2018-WichersVEL #predict #video

Hierarchical Long-term Video Prediction without Supervision (NW, RV, DE, HL), pp. 6033–6041.