Keeping Neural Networks Simple by Minimizing the Description Length of the Weights

van Camp

pages

9

• paper