Recent Progress in Algorithmic Robust Statistics via the Sum-of-Squares Method

Explore recent advancements in algorithmic robust statistics through the lens of the Sum-of-Squares method in this 51-minute lecture by Pravesh Kothari from Carnegie Mellon University. Delve into the field of learning and testing in high dimensions, covering key topics such as introduction to the subject, context, models, goals, and the blueprint for robust statistical algorithms. Examine the Meta Algorithm and its applications, understand the concept of Certifiable Subgaussian distributions, and investigate an open question in the field. Learn about list decodable regression and its implications for robust statistics. Gain valuable insights into the current state and future directions of this rapidly evolving area of computer science and statistics.

Introduction
Context
Models
Goals
Blueprint
Meta Algorithm
Certifiable Subgaussian
Open Question 1
List decodable regression
Conclusions