Abstract:
Contemporary problems in statistics and data science almost always involve solving an optimization or computation problem. Therefore, effective computational algorithms are of crucial importance in both practical and theoretical world. This workshop brings together experts who have done pathbreaking work in modern computational statistics. The objectives of the workshop include discussing recent developments in nonconvex optimization, distributed optimization, MCMC, variational methods and more; identifying important problems and new research directions, stimulating interest and fostering collaborations among young researchers.
Program Information
Guest Speakers
Yuejie Chi, Professor, Department of Electrical & Computer Engineering, Carnegie Mellon University, (virtual speaker), “Offline Reinforcement Learning: Towards Optimal Sample Complexity and Distributional Robustness”
Matt Hoffman, Research Scientist, Google,(virtual speaker), “Running Many-Chain MCMC on Cheap GPUs”
Yufeng Liu, University of North Carolina, (virtual speaker), “Distribution-Free Contextual Dynamic Pricing”
Rajesh Ranganath, Assistant Professor, Computer and Data Science, New York University, “Out of Distribution Generalization”
Max Simchowitz, Postdoctoral Researcher, CSAIL, Massachusetts Institute of Technology, “Randomized Smoothing, Online Learning, and Planning Through Contact”
Ambuj Tewari, Professor, Department of Statistics and Department of EECS, University of Michigan, (virtual speaker), “The Tortoise and the Hare: How Learning Theory is Catching Up to Applied Machine Learning”
Martin Wainwright, Cecil H. Green Professor, Massachusetts Institute of Technology, “Computational‐statistical Interplay in Reinforcement Learning”
Yuting Wei, Assistant Professor, Department of Statistics & Data Science, Wharton School, University of Pennsylvania, (virtual speaker), “A Non-asymptotic Framework for Approximate Message Passing”
Hua Zhou, Professor, Department of Biostatistics, University of California, Los Angeles, (virtual speaker), “Orthogonal Trace-sum Maximization: Applications, Local Algorithms, and Global Optimality”
Hui Zou, Professor of Statistics, University of Minnesota, (virtual speaker), “High-dimensional Clustering via Latent Semiparametric Mixture Models”
Conference Video Recording
Friday, January 13, 2023-Morning Session
Friday, January 13, 2023-Afternoon Session
Saturday, January 14, 2023 Session
Organizers
Dr. Zhihua Su, Associate Professor and Workshop Chair
Dr. George Michailidis, Director of UF Informatics Institute
Dr. Aaron Molstad, Assistant Professor of Statistics
Dr. Michael Daniels, Chair of Dept. of Statistics