Learning Compact Neural Networks with Regularization

Samet Oymak

Abstract: Proper regularization is critical for speeding up training, improving generalization performance, and learning compact models that are cost efficient. We propose and analyze regularized gradient de...