<![CDATA[Grouping to get class similarity and dividing into several classes is one of the processes of data mining. The accuracy of the grouping is an important factor to determine the product's interest in the market. The purpose of this study is to determine which products are classified as the most desirable, desirable and lessdesirable markets, so that petrified in decision making. This research uses the CRISP-DM (Cross Industry Standard Process for Data Mining) method is a methodology from minimum data that is used to analyze problems in business processes or research units. K-Means algorithm is used for grouping products that are of interest to the market. K-Means algorithm partitioned class similarity based on predetermined parameters, by calculating the centroid distance in a class. This research resulted in a product determination information system that is of interest to the market. From the test results using six parameters, namely, the number of transactions, sales volume, product categories, product diversity, average sales and number of stocks with transaction data of 1,235 transactions. Obtained the three best clusters, performance testing has been done using the Elbow method with the most SSE difference of 28,00782. Keywords: Data mining, K-Means, clustering, products market demand.]]>
Copyrights © 2021
