Abstract: Stock Market (SM) prediction is an interesting financial topic that has attracted the attention of researchers for the last years. SM dataset is a chaotic, non-linear, dynamic, non-stationary, noisy in behavior and quite difficult. The aim of this study is to predict the future SM price. The proposed system consists of three major stages, the first stage is preprocessing data that focuses on preparing the data for mining process. It includes features extraction; transform nominal to numeric and interpolation. The second stage involves building new prediction model called: Deep Neural Network (DNN) Regression Predictor (RP) that it is used for price prediction. The DNN-RP consists of three layers (technical indicators generation layer, standardization regression layer and regression predictor layer). In third stage, the evaluation has been performed depending on popular measures of prediction and 10-Cross Validation (CV). The DNN-RP is evaluated by Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) measures. The system is applied on the three global datasets of SMs (Dow, S and P500 and NASDAQ) in addition to one local SM specifically is (Bank of Baghdad). Finally, the proposed system has been compared with other popular methods. The DNN-RP is compared with K-Nearest Neighbor (KNN) and Support Vector Machine (SVM)), the results of the proposed system are better for all evaluation parameters. The best MAE and RMSE values of DNN-RP have reached to 0.0051 and 0.007, respectively.