Artificial Neural Networks: A Financial Tool As Applied in the Australian Market

Abstract

Artificial Neural Networks (ANNs), which fall under the Artificial Intelligence (AI) methodology, have enjoyed a recent revival in interest associated with advances in computer technology. While ANNs were originally developed to model the biological behavior of brain cells call neurons, recent years have seen them applied in such diverse areas as detecting bombs in airports, cancer cells in pap smears, fraud in credit card transactions, schizophrenia in mental patients, heart attacks, and foreign exchange forecasting.This thesis examines the viability of applying ANNs to problems in the finance domain using Australian financial data. The research focuses on two particular areas: predicting distress in credit unions of New South Wales and modeling the Australian/US dollar foreign exchange trading systems. The thesis examines the methodologies involved in applying ANNs to these problems as well as comparing their results with those of more conventional econometric methods.

Date of Award	8 Feb 1997
Original language	English
Supervisor	Jeffrey Carmichael (Supervisor) & Ah Chung Tsoi (Supervisor)