ICML 2019

797 papers

Top Research Directions

Research directions most represented by ICML 2019 papers.

Scaling Gaussian Processes Beyond Cubic Complexity 11
Clean Accuracy Degradation in Machine Learning Defenses 10
Sample Efficient Reinforcement Learning 9
Stationary Point Convergence in Nonconvex Optimization 8
The Pareto Frontier of Algorithmic Accuracy and Fairness 8
The Width Gap Between Deep Learning Theory and Practice 7
Beyond the Deterministic Transition Assumption in Reinforcement Learning 7
Lipschitz Regularization for Stable Neural Networks 7
Multi-Agent Learning Under Bandit Feedback 7
Mitigating Gradient Variance in Policy Optimization 7
Matrix-Free Curvature Estimation for Deep Learning 7
The Non-Linear Function Approximation Gap in Reinforcement Learning 7
The Mechanics of Benign Overfitting in High Dimensions 7
Greedy Maximization for Submodular Subset Selection 6
Stochastic Noise Injection for Escaping Saddle Points 6

Top Authors

AuthorPapers
Michael Jordan7
Sergey Levine6
Volkan Cevher6
Andreas Krause5
Bernhard Schölkopf5
Masashi Sugiyama5
Pieter Abbeel5
Stefano Ermon5
Csaba Szepesvari4
Dhruv Batra4
Emma Brunskill4
James Y. Zou4
Jinwoo Shin4
Jun Zhu4
Lawrence Carin4
Michael A Osborne4
Mingsheng Long4
Peter Bailis4
Peter Bartlett4
Sham M. Kakade4

Impactful Papers

Papers from ICML 2019 that shaped research clusters.

Foundational

Influential

Papers

Title
Training CNNs with Selective Allocation of Channels
Training Neural Networks for and by Interpolation
Training Neural Networks with Local Error Signals
Training Well-Generalizing Classifiers for Fairness Metrics and Other Data-Dependent Constraints
Trajectory-Based Off-Policy Deep Reinforcement Learning
Transferability vs. Discriminability: Batch Spectral Penalization for Adversarial Domain Adaptation
Transferable Adversarial Training: A General Approach to Adapting Deep Classifiers
Transferable Clean-Label Poisoning Attacks on Deep Neural Nets
Transfer Learning for Related Reinforcement Learning Tasks via Image-to-Image Translation
Transfer of Samples in Policy Search via Multiple Importance Sampling
Trimming the $\ell_1$ Regularizer: Statistical Analysis, Optimization, and Applications to Deep Learning
Understanding and Accelerating Particle-Based Variational Inference
Understanding and Controlling Memory in Recurrent Neural Networks
Understanding and correcting pathologies in the training of learned optimizers
Understanding and Utilizing Deep Neural Networks Trained with Noisy Labels
Understanding Geometry of Encoder-Decoder CNNs
Understanding Impacts of High-Order Loss Approximations and Features in Deep Learning Interpretation
Understanding MCMC Dynamics as Flows on the Wasserstein Space
Understanding Priors in Bayesian Neural Networks at the Unit Level
Understanding the Impact of Entropy on Policy Optimization
Understanding the Origins of Bias in Word Embeddings
Uniform Convergence Rate of the Kernel Density Estimator Adaptive to Intrinsic Volume Dimension
Unifying Orthogonal Monte Carlo Methods
Unreproducible Research is Reproducible
Unsupervised Deep Learning by Neighbourhood Discovery
Unsupervised Label Noise Modeling and Loss Correction
Using Pre-Training Can Improve Model Robustness and Uncertainty
Validating Causal Inference Models via Influence Functions
Variational Annealing of GANs: A Langevin Perspective
Variational Implicit Processes
Variational Inference for sparse network reconstruction from count data
Variational Laplace Autoencoders
Variational Russian Roulette for Deep Bayesian Nonparametrics
Voronoi Boundary Classification: A High-Dimensional Geometric Approach via Weighted Monte Carlo Integration
Warm-starting Contextual Bandits: Robustly Combining Supervised and Bandit Feedback
Wasserstein Adversarial Examples via Projected Sinkhorn Iterations
Wasserstein of Wasserstein Loss for Learning Generative Models
Weak Detection of Signal in the Spiked Wigner Model.
Weakly-Supervised Temporal Localization via Occurrence Count Learning
What is the Effect of Importance Weighting in Deep Learning?
When Samples Are Strategically Selected
Which Tasks Should Be Learned Together in Multi-task Learning?
White-box vs Black-box: Bayes Optimal Strategies for Membership Inference
Why do Larger Models Generalize Better? A Theoretical Perspective via the XOR Problem
Width Provably Matters in Optimization for Deep Linear Neural Networks
Zeno: Distributed Stochastic Gradient Descent with Suspicion-based Fault-tolerance
Zero-Shot Knowledge Distillation in Deep Networks

« prev Page 16/16