SISTEM PENJUALAN PADA PD. SAHATI BERBASIS WEB MENGGUNAKAN METODE DOUBLE EXPONENTIAL SMOOTHING UNTUK PREDIKSI STOK DAN ALGORITMA APRIORI UNTUK REKOMENDASI PRODUK

Abstract

PD. Sahati is a trading business that has been running since 2016. PD. Sahati has not utilized information technology effectively in its business activities. To avoid input errors, improve marketing and help with stock handling, a stock prediction sales system was made using the double exponential smoothing method and product recommendations using the a priori method.
Data input for prediction is product sales data. Data used for predictions cannot be less than 3 months to maintain consistency in stock predictions. The test results from the double exponential smoothing method are indicated by the mean absolute error (MAE) and mean square error (MSE). The MAE value obtained is 0.375 and the MSE value obtained is 0.145. Data input for customer product recommendations in the form of customer transaction data are then searched for frequent itemsets based on customer transaction data and will be eliminated with a minimum value of support. The minimum support value is obtained from the number of transactions divided by 2. After that, the confidence percentage will be searched through the association rules table. The product that will be recommended later is a product that has a percentage of confidence above the minimum confidence value. The minimum value of confidence in the system is 70%.