Completed
Intro
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Let's Make Block Coordinate Descent Go Fast
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Why Block Coordinate Descent?
- 3 Block Coordinate Descent for Large-Scale Optimization
- 4 Why use coordinate descent?
- 5 Problems Suitable for Coordinate Descent
- 6 Cannonical Randomized BCD Algorithm
- 7 Better Block Selection Rules
- 8 Gauss-Southwell???
- 9 Fixed Blocks vs. Variable Blocks
- 10 Greedy Rules with Gradient Updates
- 11 Gauss-Southwell-Lipschitz vs. Maximum Improvement Rule
- 12 Newton-Steps and Quadratic-Norms
- 13 Gauss-Southwell-Quadratic Rule
- 14 Matrix vs. Newton Updates
- 15 Newton's Method vs. Cubic Regularization
- 16 Experiment: Multi-class Logistic Regression
- 17 Superlinear Convergence?
- 18 Optimization with Bound Constraints
- 19 Manifold Identification Property
- 20 Superlinear Convergence and Proximal-Newton
- 21 Message-Passing for Sparse Quadratics
- 22 Experiment: Sparse Quadratic Problem
- 23 Summary
