This paper illustrate the problem of predicting movement of the Companies' Social Responsibility index (S&P EGX) using historical data for 10 years in the form of daily data, applying on Artificial Neural Network (ANN) and Random Forest by using ten technical indicators as inputs to these models. This study divides S&P Index into segments by converting inputs from continuous to separate data, so separate form indicating the movement of the direction up or down based on their inherent properties. It focuses also on comparing the performance of these models in predicting when inputs are represented in real value from and specify direction of data. Where the study for both models, but Neural approved the efficiency of the classification network Model more accurate than Random forest Model.