Segmentasi Jenis Hunian di Wilayah Bandung Menggunakan Algoritma Clustering K-Means

Abstrak

The research on property segmentation in Bandung City is motivated by the rapid growth of the property sector, accompanied by market complexity due to variations in price, specifications, and location, which makes it difficult for both buyers and developers to comprehensively understand market characteristics and make informed decisions. To address this issue, this study employs a Knowledge Discovery in Database (KDD) approach, which includes data selection, cleaning, transformation, processing, and evaluation stages, by applying the K-Means Clustering algorithm to a dataset of 57,190 property records from August 2024 to January 2025. The dataset covers variables such as price, installment, land and building area, number of rooms, and location. The determination of the optimal number of clusters was validated using the Elbow Method and Silhouette Score, resulting in two main segments with a Silhouette Score of 0.335: Cluster 0 represents the regular housing segment with 32,424 units and an average price of IDR 1.23 billion, reflecting the middle and more affordable market, while Cluster 1 represents the exclusive housing segment with 24,766 units and an average price of IDR 2.97 billion, representing the upper segment with higher specifications. The results indicate that the K-Means algorithm is effective in mapping large and complex property data, providing practical benefits for buyers in selecting homes according to their preferences and purchasing power, and for developers as a basis for formulating more targeted marketing strategies based on market segmentation.