Abstract
ABSTRACT Forecasting has recently been considered one of the most important topics in the financial investment scopes. The indelible investor may desire to uncover some aspects from uncertainty that associate with the investment decision making process, which make him rely heavily on the various analytical audits for arriving to more accurate forecasts, and support the decision making process. The research focused on using and applying number of the linear models, besides applying the back propagation Neural Networks estimating and testing the accuracy of indices closing prices forecasts' in Arab capital markets, depending on the time series for period between [1/1/2004 – 16/10/2004], and finding the best model that will be a form of depending on the results of the accuracy tools. The research concluded that the back propagation Neural Networks had the tremendous power for forecasting and it overcomes all linear models in the forecasting with depending on the results of the accuracy tools, which reversed lower value for back propagation model than linear models for the same period. .