Differentiable Sparsification for Deep Neural Networks

Yognjin Lee

Deep neural networks have significantly alleviated the burden of feature engineering, but comparable efforts are now required to determine effective architectures for these network…