ADSI Summer Workshop- Algorithmic Foundations of Learning and Control, Pablo Parrilo
Offered By: Paul G. Allen School via YouTube
Course Description
Overview
Explore advanced concepts in machine learning and control theory through this 51-minute lecture from the 2019 ADSI Summer Workshop. Delve into "Better Gaussian Kernel Factorization via Approximation Theory, with Applications" presented by Pablo Parrilo from MIT. Learn about optimal transport matrix scaling, control entrywise errors, and key ideas in approximation with polynomials. Discover plane wave expansion techniques, modern quadratic factors, and harmonic polynomials. Examine the product space of features and empirical results in linear classification. Gain insights into cutting-edge algorithmic foundations that bridge learning and control systems.
Syllabus
Intro
Why do we care
Optimal transport matrix scaling
Control entrywise errors
Three key ideas
Approximation with polynomials
Computing a change of expansions
Plane wave expansion
Plane wave expansion summary
Sustainment
Modern quadratic factor
Summary
Harmonic Polynomials
Product Space of Feature
Empirical Results
Methods
Linear Classification
Conclusion
Taught by
Paul G. Allen School
Related Courses
Sparse Representations in Signal and Image Processing: FundamentalsTechnion - Israel Institute of Technology via edX Filters and Other Potions for Early Vision and Recognition
MITCBMM via YouTube A Function Space View of Overparameterized Neural Networks - Rebecca Willet, University of Chicago
Alan Turing Institute via YouTube Approximation with Deep Networks - Remi Gribonval, Inria
Alan Turing Institute via YouTube Neural Networks as Interacting Particle Systems
Alan Turing Institute via YouTube