Computer Assisted Improvement of the Estimation Mean Squared Error with Application to Back Propagation Neural Networks.
Abstract
A computer assisted method for improving the mean squared error (MSE) in estimation for parametric models is presented. Assuming existence of nontrivial sufficient statistics, the method involves generation of Monte Carlo samples from the conditional distribution of the observables, given the sufficient statistic(s). The method is illustrated in connection with a simple back-propagation neural network model for estimating a logistic regression function. Key Words and Phrases: Parametric estimation, exponential families, nonlinear models, nonlinear least squares, neural networks, Monte Carlo simulation, computer intensive statistical methods.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 24, 1991
- Accession Number
- ADA247724
Entities
People
- J. E. Angus
Organizations
- Naval Health Research Center