Stanford AI Lab Papers and Talks at AISTATS 2021

The International Conference on Artificial Intelligence and Statistics (AISTATS) 2021 is being hosted virtually from April 13th – April 15th. We’re excited to share all the work from SAIL that’s being presented, and you’ll find links to papers, videos and blogs below. Feel free to reach out to the contact authors directly to learn more about the work that’s happening at Stanford!

List of Accepted Papers

Active Online Learning with Hidden Shifting Domains

Authors: Yining Chen, Haipeng Luo, Tengyu Ma, Chicheng Zhang


Links: Paper

Keywords: online learning, active learning, domain adaptation

A Constrained Risk Inequality for General Losses

Authors: Feng Ruan


Keywords: constrained risk inequality; super-efficiency

Comparing the Value of Labeled and Unlabeled Data in Method-of-Moments Latent Variable Estimation

Authors: Mayee F. Chen, Benjamin Cohen-Wang, Stephen Mussmann, Frederic Sala, Christopher Ré


Links: Paper

Keywords: latent variable graphical model, method-of-moments, semi-supervised learning, model misspecification

Efficient computation and analysis of distributional Shapley values

Authors: Yongchan Kwon, Manuel A. Rivas, James Zou


Links: Paper | Website

Keywords: data valuation, distributional shapley value

Improving Adversarial Robustness via Unlabeled Out-of-Domain Data

Authors: Zhun Deng, Linjun Zhang, Amirata Ghorbani, James Zou


Links: Paper

Keywords: adversarial robustness, deep learning, out of domain data

Misspecification in Prediction Problems and Robustness via Improper Learning

Authors: Annie Marsden, John Duchi, Gregory Valiant


Award nominations: Oral Presentation

Links: Paper

Keywords: machine learning, probabilistic forecasting, statistical learning theory

Online Model Selection for Reinforcement Learning with Function Approximation

Authors: Jonathan Lee, Aldo Pacchiano, Vidya Muthukumar, Weihao Kong, Emma Brunskill


Links: Paper

Keywords: reinforcement learning, model selection

Right Decisions from Wrong Predictions: A Mechanism Design Alternative to Individual Calibration

Authors: Shengjia Zhao, Stefano Ermon


Award nominations: Oral

Links: Paper | Blog Post

Keywords: uncertainty, trustworthiness, reliability

We look forward to seeing you virtually at AISTATS!

Read More