Application of K-Means Algorithm for Costumer Grouping

Abstract

Sales fluctuations are risks that must be faced by business people, PT. Gunung Hijau Success experienced this in 2016, with GAP being quite high. By giving rewards to loyal customers, it is expected to stabilize sales in the next period. So the company needs customer grouping based on customer loyalty to reward. The application of data mining can be used as an analysis to determine the loyal customer inventory according to the total purchase. In the data mining method, the clustering algorithm is one of the most popular to use where the data belonging to the same cluster will be close to each other and will be far from the data about another cluster. The results obtained in the form of customer information with criteria not loyal, loyal, and very loyal based on sales data in 2016. Also, customer criteria information from the clustering process can be used as a reference to determine the reward for customers.

Computer User Behavior Anomaly Detection Based on K-Means Algorithm

“…(3) Multiple repeatability. Due to the contingency of user behavior, a certain behavior will appear multiple times within a certain period of time, and there is still a correlation and temporal relationship between these same behaviors, which increases the difficulty of sequential pattern mining [14]. For example, a user's behavior sequence is 〈I 1 , I 2 , I 1 , I 3 , I 2 , I 1 〉, and these behaviors belong to the same transaction, and behavior I 1 , I 2 occurs multiple times, but due to the time sequence relationship between the behaviors, they cannot be merged; this is exactly the problem that traditional sequential pattern mining algorithms cannot solve.…”

Discrete Event Simulation Model Performed with Data Analytics for a Call Center Optimization

“…For the customer segmentation, the disinfected customer data taken from the company contained 6,030,355 customers, and 16 attributes of each customer were used applying k-means clustering and RFM analysis. K-means methodology is one of the most common methods used for customer clustering (Figueiredo et al, 2003;Niyagas et al, 2006;Windorto et al, 2019;Maheshwari et al, 2019;Rojlertjanya, 2019;Gustriansyah et al, 2020;Mousavi et al, 2020;Nugraha, 2020). The primary purpose of the k-means clustering is to form clusters that "minimize the squared error criterion" using the predetermined number of k values, which represents the number of clusters (Ye et al, 2013).…”