Explore recent advancements in supervised learning with noise in this 58-minute lecture by Ilias Diakonikolas from UW Madison. Delve into various noise models, challenges in high-dimensional learning and testing, and key results in the field. Gain insights into the intuition behind these developments, understand the massage noise concept, and examine the proof structure. Investigate the general case, boosting techniques, and distribution-specific learning. Learn about the tobacco noise model and its associated difficulties. Conclude with a comprehensive summary of the latest progress in supervised learning with noisy data.
Recent Developments in Supervised Learning With Noise
Overview
Syllabus
Introduction
Noise Models
Difficulties
Main result
Intuition
Massage Noise
The Proof
The General Case
Summary
Boosting
Distribution Specific Learning
Tobacco Noise Model
Tobacco Noise Difficulty
Conclusion
Taught by
Simons Institute
