Muhammad Afrizal, Ilham Saputra, Riyan Satria, Rahmaddeni Rahmaddeni

Analisis Performa Algoritma K-Means Clustering untuk Segmentasi Pasar di UMKM

Introduction

Analisis performa algoritma k-means clustering untuk segmentasi pasar di umkm. Analisis performa K-Means clustering untuk segmentasi pasar di UMKM. Identifikasi klaster pelanggan berdasarkan demografi & perilaku pembelian untuk strategi pemasaran yang efektif.

Abstract

This study aims to analyze the performance of the K-Means Clustering algorithm in market segmentation for Micro, Small, and Medium Enterprises (MSMEs). Using a quantitative approach, the data collected includes demographic information, purchasing behavior, and product preferences from respondents. The analysis process begins with data preprocessing, including normalization and outlier removal, before applying the K-Means algorithm to group customers into several segments. The performance evaluation of the algorithm is conducted using the Silhouette Score and Davies-Bouldin Index metrics. The analysis results indicate that the K-Means algorithm successfully identifies four distinct customer clusters, each with unique characteristics. The average Silhouette Score of 0.72 and a Davies-Bouldin Index of 0.45 suggest that the resulting clusters are well-defined and clearly separated. These findings provide valuable insights for MSMEs in formulating more effective and targeted marketing strategies.

Review

This study, "Analisis Performa Algoritma K-Means Clustering untuk Segmentasi Pasar di UMKM," addresses a highly relevant and practical application of data science for Micro, Small, and Medium Enterprises (MSMEs). The abstract clearly outlines the objective of analyzing K-Means performance in market segmentation, a critical task for businesses seeking to optimize their marketing strategies. The quantitative approach employed to analyze demographic information, purchasing behavior, and product preferences positions this research as a valuable contribution to the understanding of how machine learning can empower MSMEs. Methodologically, the study appears robust. The abstract highlights essential data preprocessing steps, including normalization and outlier removal, which are crucial for ensuring the quality of the clustering results. The choice of K-Means is appropriate for this type of segmentation task, and the evaluation metrics, Silhouette Score and Davies-Bouldin Index, are standard and well-regarded for assessing cluster quality. The reported results are particularly compelling: the successful identification of four distinct customer clusters, supported by an average Silhouette Score of 0.72 and a Davies-Bouldin Index of 0.45, strongly suggests that the clusters are both well-defined and clearly separated, indicating effective performance of the algorithm. The practical implications of these findings are significant, offering actionable insights for MSMEs to develop more effective and targeted marketing campaigns. The clear identification of distinct customer segments can lead to more personalized communication and product development. While the abstract provides a strong summary, future work could explore the scalability of these findings across different MSME sectors or larger datasets, and potentially compare K-Means performance against other clustering algorithms to provide a broader context. Nonetheless, this research presents a valuable demonstration of data-driven decision-making for MSMEs.

Full Text

Comments

You need to be logged in to post a comment.