A Study of Discriminant Analysis and Artificial Neural Network in Prediction of Stock Market in Nigeria

摘要

This paperanalyses financial and macroeconomic data responsible for predicting stockmarket in Nigeria using four company-specific variables and five macroeconomicvariables. The variables are inflation, investment, consumer price index,unemployment, lending interest rate, net revenue, net income and net asset.Discriminant analysis and artificial neural network were employed to determinethe variables responsible for good and bad investment choices. The result of studyhas shown that earnings per share, lending interest rate, inflation and netincome are important variables that contributed towards good and poorinvestment choices and the ANN model trained with scaled conjugate gradientalgorithm using five hidden nodes perform better in discriminating between goodand poor investment choices and has higher percentages of classifying groups.