|Abstract||Predicting stock market prices is regarded as a challenging task of financial time series, due to its chaotic, non-linear, non-stationary and dynamic nature. In this project we address the problem of stock market forecasting by making a comparison between different machine learning prediction models mainly Support Vector Machine (SVM), Artificial Neural Networks (ANN), Random Forest (RS), and Long Short Term Memory (LSTM) Recurrent Neural Network. For this goal, different models are built for predicting stock prices for 10 days in advance, and a number of experiments were executed based on ten years of historical data for stock prices from different sectors of the industry of the Qatari and the American markets. The results were analyzed using Mean Squared Error (MSE) and Mean Absolute Error (MAE) measuring metrics.
Furthermore, we developed an application for predicting stock prices and trend movement with a motivation that trading strategies and investment decisions are more reliable and efficient when guided by forecasts which could lead to more profit.