Implementation of Regression Tree Algorithm for Estimating Bread Sales (Case Study: Bysea Bites Pontianak)

Abstract

This research aims to predict the sales pattern of Bysea Bites Bakery products in Pontianak, West Kalimantan, Indonesia, using sales data. By taking 1,150 samples, consisting of date (DD-MM-YYYY), product name, and quantity sold. Using training (80%) and testing (20%) subset data, the Regression Tree algorithm was applied to predict when wheat bread would experience high sales and to find the influential factors. The performance of the model was assessed using Root Mean Square Error(RMSE) and Normalized Root Mean Square Error(NRMSE). The accuracy of the model is 73.78%. The prediction shows that on the 3, 13, 19, and 21 are dates with high demand for wheat bread and high sales on Monday, Tuesday, Thursday, Friday, and Saturday. These insights allowed the bakery to optimize its inventory, and improve customer satisfaction. Future research can examine the influence of external factors on sales patterns.