Weekly summary 2017-03-18

Posted: 2017-03-18 , Modified: 2017-03-18

Tags: none

Adversarial training
Langevin Monte Carlo
Using Barron’s Theorem
Entropy regularizer for SVM
Generalization
RL

Adversarial training

Experiments

Langevin Monte Carlo

Learn background. (OV14, Villani book, Ledoux book.)
Langevin papers.
Extending [AH16].

Using Barron’s Theorem

Entropy regularizer for SVM

Generalization

RL

Come up with project. Concrete problems in AI safety?