Prediksi Harga Saham Bank Bri Menggunakan Algoritma Linear Regresion Sebagai Strategi Jual Beli Saham

Syahputra, Janur and Ramadhani, Rima Dias and Burhanudin, Auliya (2022) Prediksi Harga Saham Bank Bri Menggunakan Algoritma Linear Regresion Sebagai Strategi Jual Beli Saham. Journal of Dinda.

Text Plagiarism_Prediksi_Harga_Saham_Bank_Bri_Menggunakan_Algoritm.pdf Download (2MB)

Abstract

Shares are securities as proof of ownership of investor in a company. Stock have a volatile nature, this makes stocks difficult to predict. Stock prediction is an effort to estimate the stock price, especially in the Bank Rakyat Indonesia company that will appear in the future, and to increase investors' profit opportunities in making investment decisions. During the COVID-19 pandemic, Bank BRI's shares experienced significant ups and downs in four months, which illustrates in the sensitivity of the stock to an event. Therefore, it is important to predict stock prices to reduce the risk accepted by investors. The prediction itself requires time series data. Time series is data that is collected sequentially from time to time. The method used for time series data is Linear Regression because this method can handle time-series data. Based on these problems, stock prediction research will be conducted at the Bank Rakyat Indonesia company using the Linear Regression method. Bank Rakyat Indonesia share price data were obtained from the investing.com website from the period starting on January 1, 2008, to June 1, 2020. The data is processed starting form preprocessing to determine attributes remove unnecessary attributes, and change the contents of the data type, then process split data to divide the dataset into training and test data. The attributes used in the study are Date and Price and the distribution of the data used is 60:40, 65:35, 70:30, 75:25, and 80:20. The best ratio is at 80:20 which produces train and test accuracy of 0.89 and 0.91, then each training data and testing data are entered into the linear regression model for prediction. The error results from the predictions were calculated using MAPE and yielded a percentage of 13.751% for training data. 13.773% for test data, and 13.755% for overall data. The MAPE results indicate that the linear regression method can be used to predict the stock price of BRI Bank

Actions (login required)

View Item