Learning Sparse Neural Networks through L_0 Regularization
international conference on learning representations, 2018.
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
PPT (Upload PPT)