Application of ARIMA model and deep learning in forecasting stock price in Vietnam

Introduction: Time series forecasting is essential in production, business, and policy-making. In Vietnam, statistical models have been used to forecast time series, such as foreign investment and consumer price index. However, only some studies have used deep learning models in predicting economic variables. The study aims to application of ARIMA model and deep learning in forecasting stock price in VietnamMethods: This article aims to review and evaluate the predictive ability of the ARIMA and deep learning models when forecasting the prices of the five stocks with the largest capitalization from January 2018 to April 2023. Results: The deep learning model's prediction results show that the deviation rate between the actual and predicted values is 3.3%. Conclusions: Businesses and stakeholders should increase the use of technology and artificial intelligence for forecasting to support decision-making