ABSTRACT
Stock investors expect optimal returns on investing in the stock market, so they need ways to be able to predict the stock price that will be their portfolio. The stock price movement that is time-series, can be predicted by various theories in financial mathematics. Deep learning applies mathematical algorithms and modeling to create many neural networks for various needs and solutions, one of which can predict the movement of time-series values. The financial sector leads the LQ45 index in Indonesia Stock Exchange market; this paper will make stock price predictions in a year using the Long Short-Term Memory (LSTM) Network Algorithm and Modeling. We use stock price data for ten years from 2009 to 2018, where stock daily closing prices are used for learning data. The goal is to predict the price of stock daily closing of the financial sector list on LQ45 index for year 2018 by using a history of nine years back and then measuring the accuracy of stock price predictions using LSTM Network.
