New results on transfer learning


#1

There is an interesting blog post on a recent paper about finding best NN architectures, and their suitability for transfer learning: