Predict Buy and Sell Time for Stock Price Using Convolutional Neural Network and Long Short-Term Memory

Abstract

Investment in stock price, analysis like fundamental analysis, technical analysis, and sentiment analysis are the key to a successor generally used by the trader. Still, not often we might find someone having loss even already analysis, that things make people scared and carefully choose a stock price. In our research, we create a program that can give forecasting times for buy and sell for stock price base on Deep Learning using Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) architecture, that can help people determine or increase confidence about conclusion after analytic that you have done. After compared the experiment result of the LSTM, ANN, CNN, and CNN-LSTM model, the highest Return is LSTM, namely for train data is Rp. 4941.91 and test data Rp. 556.468.