@article{Jusman_Nur’eni_Handayani_2022, title={Ensemble K-Nearest Neighbors Method to Predict Composite Stock Price Index (CSPI) in Indonesia}, volume={18}, url={https://journal.unhas.ac.id/index.php/jmsk/article/view/19641}, DOI={10.20956/j.v18i3.19641}, abstractNote={<p>The Composite Stock Price Index (CSPI) is a guide for investors to see the movement of stock prices as a whole from time to time. These movements always change from time to time, so it is necessary to use analytical methods to make predictions. The method that can be used to examine this is the K-Nearest Neighbor method. The combination of the results of several K-NN predictions is an effective way to get one final prediction result, namely the method ensemble K-NN. The response variable used in this study is the Composite Stock Price Index (CSPI), while the predictor variables are the gold price, the rupiah exchange rate against the dollar, and the Dow Jones Industrial Average (DJIA) index. The data used are 52 periods. The data used for training are 39 periods and the data used for testing is 13 periods. The prediction results from the ensemble have better results than the K-NN. The prediction results from the ensemble have better results than the single K-NN. The prediction results from the method are ensemble K-NN average of 6078, 634 with a MAPE value of 7,16% including high accuracy</p>}, number={3}, journal={Jurnal Matematika, Statistika dan Komputasi}, author={Jusman, Moh. and Nur’eni, Nur’eni and Handayani, Lilies}, year={2022}, month={May}, pages={423-433} }