DMCA
Combined Neural Network Models for Epidemiological Data: modelling heterogeneity and reduction of input correlations
| Citations: | 1 - 0 self |
BibTeX
@MISC{Lamers_combinedneural,
author = {Maarten H. Lamers and Joost N. Kok and Erik Lebret},
title = {Combined Neural Network Models for Epidemiological Data: modelling heterogeneity and reduction of input correlations},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We consider an epidemiological dataset, concerned with predicting pulmonary response to air pollution. To gain more knowledge of nonlinear effects and interactions in the data, nonlinear neural network techniques were applied to model the data. Initially, we modelled the data with standard feedforward network models. Based on the epidemiologic effect of heterogeneity in response, we propose a novel combined neural network modelling strategy to improve prediction quality. Also, we propose the use of a neural network strategy for reducing correlation between covariants to improve modelling quality. The results presented are promising when compared to standard feedforward network modelling. 1 Introduction One way of learning to understand the cause of diseases is to search for associations between exposures to various conditions and incidence of the studied disease or effect in a population. Such associations may provide testable hypotheses of causal links between exposure and its effec...
Keyphrases
neural network model epidemiological data modelling heterogeneity input correlation studied disease feedforward network nonlinear neural network technique epidemiological dataset various condition air pollution epidemiologic effect prediction quality nonlinear effect testable hypothesis causal link standard feedforward network model pulmonary response neural network neural network strategy
