NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Collections About News Help Login
Back to Results
Markov Chain Monte Carlo Bayesian Learning for Neural NetworksConventional training methods for neural networks involve starting al a random location in the solution space of the network weights, navigating an error hyper surface to reach a minimum, and sometime stochastic based techniques (e.g., genetic algorithms) to avoid entrapment in a local minimum. It is further typically necessary to preprocess the data (e.g., normalization) to keep the training algorithm on course. Conversely, Bayesian based learning is an epistemological approach concerned with formally updating the plausibility of competing candidate hypotheses thereby obtaining a posterior distribution for the network weights conditioned on the available data and a prior distribution. In this paper, we developed a powerful methodology for estimating the full residual uncertainty in network weights and therefore network predictions by using a modified Jeffery's prior combined with a Metropolis Markov Chain Monte Carlo method.
Document ID
20110012135
Acquisition Source
Langley Research Center
Document Type
Conference Paper
Authors
Goodrich, Michael S. (Old Dominion Univ. VA, United States)
Date Acquired
August 25, 2013
Publication Date
March 1, 2011
Publication Information
Publication: Selected Papers and Presentations Presented at MODSIM World 2010 Conference and Expo
Subject Category
Systems Analysis And Operations Research
Distribution Limits
Public
Copyright
Work of the US Gov. Public Use Permitted.

Related Records

IDRelationTitle20110012100Collected WorksSelected Papers and Presentations Presented at MODSIM World 2010 Conference and Expo
No Preview Available