ICML 2021 accepted paper list
Differential equations
- Neural Rough Differential Equations for Long Time Series
- STRODE: Stochastic Boundary Ordinary Differential Equation
- Opening the Blackbox: Accelerating Neural Differential Equations by Regularizing Internal Solver Heuristics
- Inferring Latent Dynamics Underlying Neural Population Activity via Neural Differential Equations
Generative models
- Generative Adversarial Networks for Markovian Temporal Dynamics: Stochastic Continuous Data Generation
- Marginalized Stochastic Natural Gradients for Black-Box Variational Inference
- Variational Auto-Regressive Gaussian Processes for Continual Learning
- Automatic variational inference with cascading flows
- Global inducing point variational posteriors for Bayesian neural networks and deep Gaussian processes
- Monte Carlo Variational Auto-Encoders
- BasisDeVAE: Interpretable Simultaneous Dimensionality Reduction and Feature-Level Clustering with Derivative-Based Variational Autoencoders
- n Signal-to-Noise Ratio Issues in Variational Inference for Deep Gaussian Processes
- Generative Particle Variational Inference via Estimation of Functional Gradients
- Simple and Effective VAE Training with Calibrated Decoders
- Variational (Gradient) Estimate of the Score Function in Energy-based Latent Variable Models
- On Energy-Based Models with Overparametrized Shallow Neural Networks
- Improved Contrastive Divergence Training of Energy-Based Models
- Conjugate Energy-Based Models
- Adversarial Purification with Score-based Generative Models
Dynamic systems
- Data-driven prediction of general Hamiltonian dynamics via learning exactly-symplectic maps
- Task-Optimal Exploration in Linear Dynamical Systems
- Better Training using Weight-Constrained Stochastic Dynamics
- Parallel and Flexible Sampling from Autoregressive Models via Langevin Dynamics
- Scalable Learning of Independent Cascade Dynamics from Partial Observations
Time Series
- Conformal prediction interval for dynamic time-series
- Necessary and sufficient conditions for causal feature selection in time series with latent common causes
- Approximation Theory of Convolutional Architectures for Time Series Modelling
- Whittle Networks: A Deep Likelihood Model for Time Series
- Active Learning of Continuous-time Bayesian Networks throughInterventions
- Learning Self-Modulating Attention in Continuous Time Space with Applications to Sequential Recommendation
- UnICORNN: A recurrent model for learning very long time dependencies
anomaly detection
- Event Outlier Detection in Continuous Time
Control
- Policy Analysis using Synthetic Controls in Continuous-Time
- Online Optimization in Games via Control Theory: Connecting Regret, Passivity and Poincaré Recurrence
- Global Convergence of Policy Gradient for Linear-Quadratic Mean-Field Control/Game in Continuous Time
- Model-based Reinforcement Learning for Continuous Control with Posterior Sampling
- Dropout: Explicit Forms and Capacity Control
- Locally Persistent Exploration in Continuous Control Tasks with Sparse Rewards
- Deep Coherent Exploration For Continuous Control
others
Continuous-time Model-based Reinforcement Learning
Training Recurrent Neural Networks via Forward Propagation Through Time
- Z-GCNETs: Time Zigzags at Graph Convolutional Networks for Time Series Forecasting
