Analisa Prediksi Harga Saham Blue Chip Lq45 Dengan Metode Data Mining Backpropagation Neural Network (Studi Kasus Di Bursa Efek Indonesia)

Abstract

This research is a research on predictive analysis of LQ45 blue chip stock price with backpropagation neural network data mining method. This study aims to determine the stock price prediction process using the backpropagation neural network method on the LQ45 blue chip stock price. This research is in training and testing data using RapidMiner tools with 80% data sharing for training data and 20% for testing data. The parameters used are training cycle of 500, learning rate of 0.01 and momentum of 0.9. The results of the training and testing of the stock prices of 5 companies in LQ45 obtained the RMSE (Root Mean Square Error) value with the best result of 11.296 and the largest error of 61.925 which indicates the backpropagation neural network method is quite good in the process of predicting stock prices. The results of this prediction can be used as a reference for stock investors in determining the right strategy to minimize mistakes in making decisions to buy or sell the desired stock.

Keywords: Data Mining, Backpropagation Algorithm, Neural Network, Rapidminer, Stock Price