Learning Sparse Neural Networks through L_0 Regularization

Christos Louizos
Christos Louizos

international conference on learning representations, 2018.

Cited by: 121|Views108
EI

Abstract:

We propose a practical method for $L_0$ norm regularization for neural networks: pruning the network during training by encouraging weights to become exactly zero. Such regularization is interesting since (1) it can greatly speed up training and inference, and (2) it can improve generalization. AIC and BIC, well-known model selection crit...More

Code:

Data:

Your rating :
0

 

Tags
Comments