Generalization Bounds for Neural Networks via Approximate Description Length

Amit Daniely, Elad Granot

Neural Information Processing Systems 

We investigate the sample complexity of networks with bounds on the magnitude of its weights.