Regularizing Neural Networks via Stochastic Branch Layers

Wonpyo Park
Wonpyo Park
Paul Hongsuck Seo
Paul Hongsuck Seo

ACML, pp. 678-693, 2019.

Cited by: 0|Views59
EI

Abstract:

We introduce a novel stochastic regularization technique for deep neural networks, which decomposes a layer into multiple branches with different parameters and merges stochastically sampled combinations of the outputs from the branches during training. Since the factorized branches can collapse into a single branch through a linear ope...More

Code:

Data:

Your rating :
0

 

Tags
Comments