# AISTATS 2021 Oral Sessions Schedule

All times are in Pacific Daylight Time (PDT).## Click to expand schedule for Day 1 Oral Sessions (Tue, April 13)

Session Title |
Title |
Session |

Theory of Statistical and Deep Learning Methods | Homeomorphic-Invariance of EM: Non-Asymptotic Convergence in KL Divergence for Exponential Families via Mirror Descent | Session 1: April 13 at 10:30am-11:30am PDT |
---|---|---|

Recovery Guarantees for Kernel-based Clustering under Non-parametric Mixture Models | Session 1: April 13 at 10:30am-11:30am PDT | |

Towards a Theoretical Understanding of the Robustness of Variational Autoencoders | Session 1: April 13 at 10:30am-11:30am PDT | |

Stable ResNet | Session 1: April 13 at 10:30am-11:30am PDT | |

Sampling Methods | Couplings for Multinomial Hamiltonian Monte Carlo | Session 2: April 13 at 11:30am-12:30pm PDT |

An Adaptive-MCMC Scheme for Setting Trajectory Lengths in Hamiltonian Monte Carlo | Session 2: April 13 at 11:30am-12:30pm PDT | |

Maximal Couplings of the Metropolis-Hastings Algorithm | Session 2: April 13 at 11:30am-12:30pm PDT | |

GANs with Conditional Independence Graphs: On Subadditivity of Probability Divergences | Session 2: April 13 at 11:30am-12:30pm PDT | |

Bandits, Reinforcement Learning / Optimization | Federated Multi-armed Bandits with Personalization | Session 3: April 13 at 16:15-17:15 PDT |

Near-Optimal Provable Uniform Convergence in Offline Policy Evaluation for Reinforcement Learning | Session 3: April 13 at 16:15-17:15 PDT | |

Provably Efficient Safe Exploration via Primal-Dual Policy Optimization | Session 3: April 13 at 16:15-17:15 PDT | |

Bayesian Coresets: Revisiting the Nonconvex Optimization Perspective | Session 3: April 13 at 16:15-17:15 PDT | |

Theory and Practice of Machine Learning | Entropy Partial Transport with Tree Metrics: Theory and Practice | Session 4: April 13 at 17:15-18:15 PDT |

Independent Innovation Analysis for Nonlinear Vector Autoregressive Process | Session 4: April 13 at 17:15-18:15 PDT | |

Beyond Marginal Uncertainty: How Accurately can Bayesian Regression Models Estimate Posterior Predictive Correlations? | Session 4: April 13 at 17:15-18:15 PDT | |

A Variational Information Bottleneck Approach to Multi-Omics Data Integration | Session 4: April 13 at 17:15-18:15 PDT |

## Click to expand schedule for Day 2 Oral Sessions (Wed, April 14)

Session Title |
Title |
Session |

Theory and Methods of Learning | Neural Enhanced Belief Propagation on Factor Graphs | Session 5: April 14 at 08:15am-9:15am PDT |
---|---|---|

An Analysis of LIME for Text Data | Session 5: April 14 at 08:15am-9:15am PDT | |

Bandit algorithms: Letting go of logarithmic regret for statistical robustness | Session 5: April 14 at 08:15am-9:15am PDT | |

The Sample Complexity of Level Set Approximation | Session 5: April 14 at 08:15am-9:15am PDT | |

Bandits, Reinforcement Learning / Learning Theory / Sparse Methods | Logistic Q-Learning | Session 6: April 14 at 9:15am-10:15am PDT |

Instance-Wise Minimax-Optimal Algorithms for Logistic Bandits | Session 6: April 14 at 9:15am-10:15am PDT | |

Robust and Private Learning of Halfspaces | Session 6: April 14 at 9:15am-10:15am PDT | |

Hadamard Wirtinger Flow for Sparse Phase Retrieval | Session 6: April 14 at 9:15am-10:15am PDT | |

Optimization / Learning Theory / Generalization | Projection-Free Optimization on Uniformly Convex Sets | Session 7: April 14 at 10:30am-11:30am PDT |

Measure Transport with Kernel Stein Discrepancy | Session 7: April 14 at 10:30am-11:30am PDT | |

Learn to Expect the Unexpected: Probably Approximately Correct Domain Generalization | Session 7: April 14 at 10:30am-11:30am PDT | |

Improving Adversarial Robustness via Unlabeled Out-of-Domain Data | Session 7: April 14 at 10:30am-11:30am PDT | |

Graphs and Networks | Graph Community Detection from Coarse Measurements: Recovery Conditions for the Coarsened Weighted Stochastic Block Model | Session 8: April 14 at 11:30am-12:30pm PDT |

MatÃ©rn Gaussian Processes on Graphs | Session 8: April 14 at 11:30am-12:30pm PDT | |

Differentially Private Analysis on Graph Streams | Session 8: April 14 at 11:30am-12:30pm PDT | |

On Learning Continuous Pairwise Markov Random Fields | Session 8: April 14 at 11:30am-12:30pm PDT |

## Click to expand schedule for Day 3 Oral Sessions (Thu, April 15)

Session Title |
Title |
Session |

Fairness / Privacy / Decision Making / Data Cleaning | Private optimization without constraint violations | Session 9: April 15 at 12:00-13:00 PDT |
---|---|---|

Learning Smooth and Fair Representations | Session 9: April 15 at 12:00-13:00 PDT | |

Right Decisions from Wrong Predictions: A Mechanism Design Alternative to Individual Calibration | Session 9: April 15 at 12:00-13:00 PDT | |

PClean: Bayesian Data Cleaning at Scale via Domain-Specific Probabilistic Programming | Session 9: April 15 at 12:00-13:00 PDT | |

Generalization / Reinforcement Learning / Optimization | Off-policy Evaluation in Infinite-Horizon Reinforcement Learning with Latent Confounders | Session 10: April 15 at 13:00-14:00 PDT |

Does Invariant Risk Minimization Capture Invariance? | Session 10: April 15 at 13:00-14:00 PDT | |

Density of States Estimation for Out of Distribution Detection | Session 10: April 15 at 13:00-14:00 PDT | |

Quick Streaming Algorithms for Maximization of Monotone Submodular Functions in Linear Time | Session 10: April 15 at 13:00-14:00 PDT | |

Deep Learning / High-dimensionality | Sketch based Memory for Neural Networks | Session 11: April 15 at 14:15-15:15 PDT |

Associative Convolutional Layers | Session 11: April 15 at 14:15-15:15 PDT | |

Deep Fourier kernel for self-attentive point processes | Session 11: April 15 at 14:15-15:15 PDT | |

Uniform consistency of cross-validation estimators for high-dimensional ridge regression | Session 11: April 15 at 14:15-15:15 PDT | |

Learning Theory | A constrained risk inequality for general losses | Session 12: April 15 at 15:15-16:15 PDT |

Misspecification in Prediction Problems and Robustness via Improper Learning | Session 12: April 15 at 15:15-16:15 PDT | |

Minimax Optimal Regression over Sobolev Spaces via Laplacian Regularization on Neighborhood Graphs | Session 12: April 15 at 15:15-16:15 PDT | |

Faster Kernel Interpolation for Gaussian Processes | Session 12: April 15 at 15:15-16:15 PDT |