A probabilistic framework for nonlinearities in stochastic neural networks

Published

Conference Paper

© 2017 Neural information processing systems foundation. All rights reserved. We present a probabilistic framework for nonlinearities, based on doubly truncated Gaussian distributions. By setting the truncation points appropriately, we are able to generate various types of nonlinearities within a unified framework, including sigmoid, tanh and ReLU, the most commonly used nonlinearities in neural networks. The framework readily integrates into existing stochastic neural networks (with hidden units characterized as random variables), allowing one for the first time to learn the nonlinearities alongside model weights in these networks. Extensive experiments demonstrate the performance improvements brought about by the proposed framework when integrated with the restricted Boltzmann machine (RBM), temporal RBM and the truncated Gaussian graphical model (TGGM).

Duke Authors

Cited Authors

  • Su, Q; Liao, X; Carin, L

Published Date

  • January 1, 2017

Published In

Volume / Issue

  • 2017-December /

Start / End Page

  • 4487 - 4496

International Standard Serial Number (ISSN)

  • 1049-5258

Citation Source

  • Scopus