Representation Costs of Linear Neural Networks: Analysis and Design

Channel:
United States Simons Institute for the Theory of Computing
Subscribers:
68,700
Published on ● Video Link: https://www.youtube.com/watch?v=BYGixdvjuaM


Duration: 35:50
918 views
11

Mina Karzand (University of California, Davis)
https://simons.berkeley.edu/talks/tba-154
Deep Learning Theory Symposium



Other Videos By Simons Institute for the Theory of Computing


2022-01-19Introduction to Causal Graphical Models: Graphs, d-separation, do-calculus
2022-01-19Introduction to Causal Graphical Models: Graphs, d-separation, do-calculus
2022-01-19Introduction to Causal Graphical Models: Graphs, d-separation, do-calculus
2022-01-18Introduction to Causal Graphical Models: Graphs, d-separation, do-calculus
2021-12-15Program Organizer Presentations
2021-12-15Lightning Talks by Research Fellows
2021-12-15Welcome
2021-12-15Lightning Talks by Industry Partners
2021-12-11Foundations for a Democratic Metaverse
2021-12-09The Thresholding Scheme for Mean Curvature Flow and de Giorgi’s Ideas for Gradient Flows
2021-12-08Representation Costs of Linear Neural Networks: Analysis and Design
2021-12-08On the Cryptographic Hardness of Learning One-Hidden Layer Neural Networks
2021-12-08Is Overfitting Actually Benign? On the Consistency of Interpolating Methods
2021-12-08Universality of Neural Networks
2021-12-08The Spectrum of Nonlinear Random Matrices for Ultra-Wide Neural Networks
2021-12-08Exact Asymptotics and Universality for Gradient Flows and Empirical Risk Minimizers
2021-12-07Learning Staircases
2021-12-07Sharp Matrix Concentration
2021-12-07Non-Parametric Convergence Rates for Plain Vanilla Stochastic Gradient Descent
2021-12-07Self-Training Converts Weak Learners to Strong Learners in Mixture Models
2021-12-07Graphon Neural Networks and the Transferability of Graph Neural Networks


Tags:
Simons Institute
theoretical computer science
UC Berkeley
Computer Science
Theory of Computation
Theory of Computing
Deep Learning Theory Symposium
Mina Karzand