JAKARTA COMPOSITE INDEX (JCI) PREDICTION USING K-NEAREST NEIGHBORS REGRESSION

Najibullah
Azlina
Ricky Ariansyah
Lisnawati

Jurnal Kreasi Rakyat
Vol. 2 No. 1 (2024): in February

This paper aims to evaluate the accuracy of the k-Nearest Neighbors regression method in predicting the Jakarta Composite Index (JCI), thus adding evidence of the application of k-NN regression in predicting noisy financial time series data. The dataset consists of daily data from several time series variables in 2022. The result suggests that the accuracy of k-NN regression in predicting JCI fluctuation is very good, indicated by Mean Absolute Percentage Error (MAPE) being less than 2% for k = 2, 3, …, 10.