International Journal of Computer and Information Technology (ISSN: 2279 – 0764) Volume 10 – Issue 5, September 2021 www.ijcit.com 209 Deep Learning Methods In Predicting Indonesia Composite Stock Price Index (IHSG) Arief Fadhlurrahman Rasyid, Dewi Agushinta R.*, Dharma Tintri Ediraras Magister of Information System Management, Information System, Economy Gunadarma University Depok, West Java, Indonesia * Corresponding author’s email: dewier [AT] staff.gunadarma.ac.id Abstract— The stock price changes at any time within seconds. The stock price is a time series data. Thus, it is necessary to have the best analysis model in predicting the stock price to make decisions to avoid losses in investing. In this research, the method used two models Deep Learning namely Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) in predicting Indonesia Composite Stock Price Index (IHSG). The dataset used is historical data from the Jakarta Composite Index (^JKSE) stock price in 2013-2020 obtained through Yahoo Finance. The results suggest that Deep learning methods with LSTM and GRU models can predict Indonesia Composite Stock Price Index (IHSG). Based on the test results obtained RMSE value of 71.28959454502723 with an accuracy rate of 92.39% for LSTM models and obtained RMSE value of 70.61870739073838 with an accuracy rate of 96.77% on GRU models. Keywords- Composite Stock Price Index (IHSG); Deep learning; GRU; Jakarta Composite Index (^JKSE); LSTM; predicting; stock price I. INTRODUCTION Indonesia’s investment and trading are growing quite rapidly, investing in stocks, forex or commodities is very tempting because we can get a considerable profit. Tens, even hundreds of percent in a short period (a day, a week, a month, depending on the conditions) we can get. The profit obtained can also be several times the profit of deposits. However, if there is a miscalculation, it can bankrupt investors in a short time as well. In businessinsider.com, a stock is an investment that represents a unit of ownership in a company. Stocks are securities that prove an investor's ownership of a company. This is interpreted if a person buys stocks of a company. The number of stocks purchased has given capital to the company. The Company has the option to issue stocks to help fund the company. Nevertheless, stocks are a means of investment that has a high interest among investors because stocks can provide an attractive level of profit. Every investor who invests in a company usually has the purpose to get capital gain, namely the difference between the purchase price and the selling price of stocks and dividends in the form of company profits given to investors [1]. Stocks are the easiest asset to liquidate, when we need cash, we can get it quickly. Due to the momentary valuation by both buyers and sellers influenced by several factors, the stock price changes at any time within seconds. Indonesia Stock Exchange (IDX) is one of the best-growing stock markets. The Indonesia Stock Exchange (IDX) is the capital market in Indonesia that has existed long before the Independence of Indonesia, and the first stock exchange in Indonesia was established in 1912. The Composite Stock Price Index (IHSG) is a tool to measure the performance of the Indonesian stock market. Many factors can influence the movement of the Composite Stock Price Index (IHSG), one of which is macroeconomic and global economic conditions. Macroeconomic conditions may affect the movement of IHSG, such as the occurrence of inflation and exchange rates. Meanwhile, the global economy that can affect IHSG is the world gold price and the world oil price [2]. The stock price is a time-series data that has a large dimension and is not static, so developed an approach for non- linear data [3]. Popular theories indicate that stock markets are primarily a chance step, with certain rules on the preceding day's closing price. The majority of traditional time-series forecasting techniques are based on stationary trends. As a result, stock price forecasting is fraught with difficulty. Furthermore, due to the large number of variables involved, predicting stock prices is a difficult task in and of itself. The market behaves like a voting machine in the near term, but like a weighing machine in the end, therefore there is potential for forecasting market movements over a longer timescale [4]. The volatility of the stock market is complex and nonlinear. Relying only on a trader's personal experience and intuition for analysis and judgment is unreliable and inefficient. To lead stock trading, people require an intelligent, scientific, and successful study strategy. The application of deep learning in predicting stock prices has become a research hotspot with the rapid growth of artificial intelligence. Because of its good non- linear approximation abilities and adaptive self-learning, the