Overview
Explore the fascinating intersections of physics and deep learning in this 52-minute lecture by MIT professor Max Tegmark. Delve into topics such as feedforward neural networks, AlphaGo, expressability, and polynomials, while uncovering intriguing connections between these seemingly disparate fields. Examine scale-invariant behavior, long-range correlations in magnets, and supervised learning techniques. Gain insights into the Venn diagram of physics and deep learning, and discover how these disciplines inform and enhance each other through engaging examples and thought-provoking visualizations.
Syllabus
Connections between physics and deep learning
Venn diagram
Feedforward neural nets
Alphago
Expressability
Example
Polynomials
Fun connection
Weird plot
Magnets
Longrange correlations
Scaleinvariant behavior
Supervised learning
Taught by
MITCBMM