PENERAPAN FORECASTING METHODS UNTUK PENJUALAN PRODUK UMKM DENGAN ALGORITMA K-NEAREST NEIGHBOR

Abstract

This research will be carried out on MSMEs in the city of Batam, precisely in the Cipta Asri housing phase II block fir, the analysis is an experiment on sales data so far on MSMEs engaged in the culinary field with the aim of predicting the level of sales in the future. This study will utilize the Forecasting method with the K-Nearest Neighbor datamining algorithm technique to forecast product sales and to test the suitability of the researcher's accuracy using rapidminer software. The results obtained are based on the analysis carried out from 26 training data, there are 20 data that have been classified correctly and 6 data that have not been classified correctly, namely the percentage for Correctly Classified Instances is 77.00% while the percentage for Incorrectly Classified Instances is 23.00%. In conclusion, the results of the analysis can be effective in determining the results of future sales targets and will affect the next level of sales.