Completed
Subgradient descent for machine learning Assumptions is the expected risk, the empirical risk
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Large Scale Machine Learning and Convex Optimization - Lecture 3
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Main motivating examples
- 3 Subgradient method/descent (Shor et al., 1985)
- 4 Subgradient descent for machine learning Assumptions is the expected risk, the empirical risk
- 5 Summary: minimizing convex functions
- 6 Relationship to online learning
- 7 Stochastic subgradient "descent" /method
- 8 Convex stochastic approximation Existing work • Known global minimax rates of convergence for non-smooth problems (Nemirovsky and Yudin, 1983; Agarwal et al., 2012)
- 9 Robustness to wrong constants for = Cn
- 10 Robustness to lack of strong convexity
- 11 Beyond stochastic gradient method
- 12 Outline
- 13 Adaptive algorithm for logistic regression
- 14 Self-concordance
- 15 Least-mean-square algorithm
- 16 Markov chain interpretation of constant step sizes
- 17 Least-squares - Proof technique
- 18 Simulations - synthetic examples