Prediction of the Stock Market Using LSTM, ARIMA, and Hybrid of LSTM-ARIMA Models

Abstract

This paper will be presented a hybrid ARIMA-LSTM model   for forecasting the stock price of Saudi Basic Industries Corporation (SABIC) on the TASI index, highlighting the growing importance of machine learning in financial market predictions. The advantages of the autoregressive integrated moving average (ARIMA) model are its ability to capture linear trends, and the advantages of the Long Short-Term Memory (LSTM) is good at modeling complex nonlinear relationships in time series data, so it is proposed to combine them in a hybrid model. The forecasting process begins with ARIMA, which identifies and addresses the linear components of the data. The residuals generated by ARIMA, which exhibit non-linear and stochastic characteristics, are then passed to the LSTM network to capture intricate patterns. By integrating both models, the ARIMA-LSTM hybrid approach is able to address both the linear and non-linear aspects of the stock data, leading to improved prediction accuracy. The anticipated outcomes are shown that the hybrid model outperforms individual models, which leads to the emphasizing its potential as a powerful forecasting tool. its potential as a powerful forecasting tool. This methodology is especially valuable in financial market analysis, where precise asset performance predictions are crucial for effective portfolio management and investment decisions.