Application of artificial neural networks in the prediction of earnings

Abstract

The feasibility of using artificial neural networks (ANNs) for predicting future earnings by stock market and capital market investors was evaluated. A multilayer perceptron feedforward neural network architecture with an efficient and robust training algorithm was used in the study because of its suitability as a classifier in accounting and business processes. The logistic regression method was also examined in parallel with the ANN approach to compare the performance of the two methods.