Comparison of Autoregressive Integrated Moving Average (ARIMA) model and Gated Recurrent Unit (GRU) model in predicting stock prices

Abstract

Investing in the stock market is not easy, because the stock price always fluctuates. These fluctuations can lead to losses if the prediction is wrong. Therefore, we need a method that can predict the direction of stock movements. This paper presents a comparative study for stock price prediction using Autoregressive Integrated Moving Average (ARIMA) model and Gated Recurrent Unit (GRU) model. For building and testing the models, the historical values of PT Bank Central Asia Tbk (IDX: BBCA) stock closing price and PT Kertas Tjiwi Kimia Tbk (IDX: TKIM) stock closing price for the period 01-January-2015 to 01-January-2020, were used. The result shows that the Gated Recurrent Unit (GRU) model has better accuracy than the Autoregressive Integrated Moving Average (ARIMA) model.