<![CDATA[Optimization is a process of solving a problem so that it can provide the best conditions that can provide a maximum or minimum value. In a business, optimization of stock procurement is an important thing, including in terms of product sales. If the stock of a product is empty then the sales potential decreases. Therefore we need a method to optimize the stock so that it can supply consumer demand and ultimately increase sales. Data mining can be applied in the sales system by creating a sales classification model for the best-selling products. In this study, the sales classification of the best-selling products at electronic stores was carried out using Naïve Bayes. The data used in this study is data on sales of electronic products for 3 months. In the early stages, preprocessing is carried out, namely by encoding labels. Model testing was carried out using percentage split and cross validation with several trials. Through the use of percentage split, the best accuracy is obtained at 93.3% with a comparison of 30% of test data and 70% of training data. The best accuracy using cross validation was obtained by 84% for 7-fold. The classification system that has been created is capable of classifying the best-selling products every quarter of a year. Through the use of the best-selling product classification system, the store can find out the best-selling product stock so that the stock is not empty. Thus the procurement of store stock can be more optimal and sales will increase.]]>
