USING ARTIFICIAL NETWORK FOR ASSESSING EQUITY INVESTMENT IN BANKS

Abstract

In assessing a company for the purpose of a possible investment, investment ratios are used to detect an improving or declingin trend, which can be useful to shareholders and the would-be investors.  This paper shows the application of artificial neural network for a dichotomous classification problem in assessing equity investments in banks.  Kohonen map is the neural network model used.  The performance of hohonen map is highly evaluated by training the network with a set of five financial ratios data collected from different banks organisations that are either viable or not viable.  The already trained network is then taken through a viability test by some test set of financial ratios to determine whether a bank is vaiable or not so that an individual or a corporate existence will be able to determine whether to invest in such a bank or not.    The results also show the degree of viability of a bank provided by a pie chart.