Application of Neural Network in Financial Market

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
To be successful in a competitive world, any organization must not only maintain a good understanding of the current conditions, but be able to forecast the future as accurately and precisely as possible. Analysis of historical trends in data, with a view to making predictions is therefore a key task, and one for which neural network technology is well suited. In general, application of neural network in financial market is discussed. Specifically, the prediction of foreign exchange rate using neural network called “feedforward” neural network with one hidden layer consisting of three neurons is considered in deriving the model for the forecasting. Statistical evaluation of the output from the implementation of the neural network model on the forecast values with the actual of Banks-Indicative Opening U.S. Dollar, Pound Sterling and Euro Rates – Selling, are discussed. Further comparison of the results obtained from the improved model and the existing model was considered. The results obtained suggest that neural network model produces better accurate predictions than other time series forecasting models. The use of the system in financial market and specifically on the prediction of exchange rates typically involves the utilization of other financial indicators and domain knowledge since real life data were used as input.
A Thesis submitted to the Department of Computer Engineering,Kwame Nkrumah University of Science and Technology in partial fulfillment of the requirements for the degree of Master of Philosophy, 2011