ICML 2020

1,099 papers

Top Research Directions

Research directions most represented by ICML 2020 papers.

Bypassing the Computational Bottleneck of Adversarial Training 13
Overcoming the Convexity Constraint in Modern Optimization 12
Mitigating Gradient Variance in Stochastic Optimization 11
Scaling Gaussian Processes Beyond Cubic Complexity 11
Clean Accuracy Degradation in Machine Learning Defenses 11
Mitigating Gradient Variance in Policy Optimization 10
Safe Reinforcement Learning via Constrained Markov Decision Processes 9
The Non-Linear Function Approximation Gap in Reinforcement Learning 9
Scaling Reinforcement Learning to High-Dimensional Visual Observations 8
Accelerating Neural Architecture Search Without Full Model Training 8
Navigating Non-Convex Loss Landscapes in Deep Learning 8
The Function Approximation Gap in Reinforcement Learning Theory 8
Understanding Gradient Masking in Adversarial Defenses 8
Mitigating Input Perturbation Sensitivity in Deep Networks 8
Causal Inference with Unobserved Confounding 8

Top Authors

AuthorPapers
Masashi Sugiyama11
Michael Jordan8
Dale Schuurmans7
Gang Niu7
Michal Valko7
Percy Liang7
Tommi Jaakkola7
Zhaoran Wang7
Mihaela van der Schaar6
Pieter Abbeel6
Stefano Ermon6
Yoshua Bengio6
Zhangyang Wang6
Zhuoran Yang6
Aldo Pacchiano5
Bo Han5
Chi Jin5
Heng Huang5
Jinwoo Shin5
Mehryar Mohri5

Impactful Papers

Papers from ICML 2020 that shaped research clusters.

Foundational

Influential

Papers

Title
Retro*: Learning Retrosynthetic Planning with Neural Guided A* Search
Reverse-engineering deep ReLU networks
Revisiting Fundamentals of Experience Replay
Revisiting Spatial Invariance with Low-Rank Local Connectivity
Revisiting Training Strategies and Generalization Performance in Deep Metric Learning
RIFLE: Backpropagation in Depth for Deep Transfer Learning through Re-Initializing the Fully-connected LayEr
Rigging the Lottery: Making All Tickets Winners
Robust and Stable Black Box Explanations
Robust Bayesian Classification Using An Optimistic Score Ratio
Robust Graph Representation Learning via Neural Sparsification
Robustifying Sequential Neural Processes
Robust Learning with the Hilbert-Schmidt Independence Criterion
Robustness to Programmable String Transformations via Augmented Abstract Training
Robustness to Spurious Correlations via Human Annotations
Robust One-Bit Recovery via ReLU Generative Networks: Near-Optimal Statistical Rate and Global Landscape Analysis
Robust Outlier Arm Identification
Robust Pricing in Dynamic Mechanism Design
ROMA: Multi-Agent Reinforcement Learning with Emergent Roles
Safe Deep Semi-Supervised Learning for Unseen-Class Unlabeled Data
Safe Imitation Learning via Fast Bayesian Reward Inference from Preferences
Safe Reinforcement Learning in Constrained Markov Decision Processes
Safe screening rules for L0-regression from Perspective Relaxations
Sample Amplification: Increasing Dataset Size even when Learning is Impossible
Sample Complexity Bounds for 1-bit Compressive Sensing and Binary Stable Embeddings with Generative Priors
Sample Factory: Egocentric 3D Control from Pixels at 100000 FPS with Asynchronous Reinforcement Learning
SCAFFOLD: Stochastic Controlled Averaging for Federated Learning
Scalable and Efficient Comparison-based Search without Features
Scalable Deep Generative Modeling for Sparse Graphs
Scalable Differentiable Physics for Learning and Control
Scalable Differential Privacy with Certified Robustness in Adversarial Learning
Scalable Exact Inference in Multi-Output Gaussian Processes
Scalable Gaussian Process Separation for Kernels with a Non-Stationary Phase
Scalable Identification of Partially Observed Systems with Certainty-Equivalent EM
Scalable Nearest Neighbor Search for Optimal Transport
Scaling up Hybrid Probabilistic Inference with Logical and Arithmetic Constraints via Message Passing
Schatten Norms in Matrix Streams: Hello Sparsity, Goodbye Dimension
SDE-Net: Equipping Deep Neural Networks with Uncertainty Estimates
Searching to Exploit Memorization Effect in Learning with Noisy Labels
Second-Order Provable Defenses against Adversarial Attacks
Selective Dyna-Style Planning Under Limited Model Capacity
Self-Attentive Associative Memory
Self-Attentive Hawkes Process
Self-Modulating Nonparametric Event-Tensor Factorization
Self-PU: Self Boosted and Calibrated Positive-Unlabeled Training
Self-supervised Label Augmentation via Input Transformations
Semiparametric Nonlinear Bipartite Graph Representation Learning with Provable Guarantees
Semismooth Newton Algorithm for Efficient Projections onto $\ell_1, ∞$-norm Ball
Semismooth Newton Algorithm for Efficient Projections onto $\ell_{1, \infty}$-norm Ball
Semismooth Newton Algorithm for Efficient Projections onto ℓ1, ∞-norm Ball
Semi-Supervised Learning with Normalizing Flows

« prev Page 18/22 next »