Nonconvex Optimization for High-dimensional Learning: From ReLUs to Submodular Maximization

Channel:
United States Simons Institute for the Theory of Computing
Subscribers:
68,700
Published on ● Video Link: https://www.youtube.com/watch?v=5DQ2y5HRvhI


Duration: 34:41
2,587 views
35

Mahdi Soltanolkotabi, University of Southern California
https://simons.berkeley.edu/talks/mahdi-soltanolkotabi-10-05-17
Fast Iterative Methods in Optimization



Other Videos By Simons Institute for the Theory of Computing


2017-11-06Bounds for Matrix Factorization Ranks via Semidefinite Programming and...
2017-11-06A Lower Bound on the Positive Semidefinite Rank of Convex Bodies
2017-11-06Computing the Nucleolus of Weighted Cooperative Matching Games in Polynomial Time
2017-11-06Small (Explicit) Extended Formulation for Knapsack Cover Inequalities from Monotone Circuits
2017-11-06Constructing Extended Formulations
2017-11-01Lower Bounds on the Size of Linear Programs
2017-10-27Black Holes, Firewalls, and the Limits of Quantum Computers
2017-10-06Natasha 2: Faster Non-convex Optimization Than SGD
2017-10-06From Minimum Cut to Submodular Minimization: Leveraging the Decomposable Structure
2017-10-06Will Vanishing Gradients Ever Vanish from Deep Learning?
2017-10-05Nonconvex Optimization for High-dimensional Learning: From ReLUs to Submodular Maximization
2017-10-05Faster Algorithms and New Iterative Methods for Computing the Stationary Distribution
2017-10-05Trends in Large-scale Nonconvex Optimization
2017-10-04Sketchy Decisions: Convex Low-Rank Matrix Optimization with Optimal Storage
2017-10-04Randomized Iterative Methods and Complexity for Markov Decision Process
2017-10-04Hashing-based-estimators for Kernel Density in High Dimensions
2017-10-04Sublinear Time Low-rank Approximation of Positive Semidefinite Matrices
2017-10-04Sketching as a Tool for Fast Algorithm Design
2017-10-03Dealing with Linear Constraints via Random Permutation
2017-10-03Zero-order and Dynamic Sampling Methods for Nonlinear Optimization
2017-10-03Algorithmic Tools for Smooth Nonconvex Optimization


Tags:
Ruoyu Sun
Simons Institute
Theory of Computing
Theory of Computation
Theoretical Computer Science
Computer Science
UC Berkeley
Fast Iterative Methods in Optimization