A Neural Network Approach to Financial Forecasting

Abstract

the world economy keeps on changing, financial institutions and investors always look forward to a system by which they can monitor the dynamic financial state of the world. This calls for a system that could simulate and predict financial positions based on financial market trends in order to manage and identify the best package to invest in. The paper is aimed at developing a neural network application to predict interest rate on loan investment in Nigerian bank using the back propagation neural network.It forecastinterest rate on loan investment in three areas which include commerce, education, and rent/housing. The simulation was done using Matlab 2008. From the results obtained the Mean Squared Error values 3.99104e-6 in the Training, 3.597228e-5 in the validation and 9.9464314e-6 in the testing which shows that the prediction was done with minimum amount of error. Keywordseural Network, back propagation, Mean Square Error, Training, Validation, and Prediction