Year: 2002
Type: dissertation
Abstract: In this thesis the two disciplines of Statistics and Artificial Neural Networks \nare combined into an integrated study of a data set of a weather modification \nExperiment. \nAn extensive literature study on artificial neural network methodology has \nrevealed the strongly interdisciplinary nature of the research and the applications \nin this field. \nAn artificial neural networks are becoming increasingly popular with data \nanalysts, statisticians are becoming more involved in the field. A recursive \nalgoritlun is developed to optimize the number of hidden nodes in a feedforward \nartificial neural network to demonstrate how existing statistical techniques \nsuch as nonlinear regression and the likelihood-ratio test can be applied in \ninnovative ways to develop and refine neural network methodology. \nThis pruning algorithm is an original contribution to the field of artificial \nneural network methodology that simplifies the process of architecture selection, \nthereby reducing the number of training sessions that is needed to find \na model that fits the data adequately. \n[n addition, a statistical model to classify weather modification data is developed \nusing both a feedforward multilayer perceptron artificial neural network \nand a discriminant analysis. The two models are compared and the effectiveness \nof applying an artificial neural network model to a relatively small \ndata set assessed. \nThe formulation of the problem, the approach that has been followed to \nsolve it and the novel modelling application all combine to make an original \ncontribution to the interdisciplinary fields of Statistics and Artificial Neural \nNetworks as well as to the discipline of meteorology.
Author: L. Royal Fletcher
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1542469369