The Implementation of Principal Component Analysis (PCA) and K-Means Clustering for Heavy Metals Content in Well Water in Bandung Regency Devri Maulana, Udjianna Sekteria Pasaribu, Nurul Fahimah, Katharina Oginawati, Kurnia Novita Sari
ITB
Abstract
Excessive levels of heavy metals in water sources can degrade water quality and harm humans. Nowadays, there has been little research on heavy metal content in well water in the Upper Citarum Watershed. Well water samples were taken from 160 locations, each containing ten heavy metals (arsenic, cadmium, chromium, cobalt, copper, iron, lead, manganese, mercury, and zinc). To determine the similarity between locations based on heavy metal content, cluster analysis was conducted using the K-Means method. However, the dimensions of the data are quite large (ten dimensions) so it is necessary to reduce the dimensions with principal component analysis (PCA) first. From the PCA results, the initial ten variables were reduced to eight new variables, capturing 87.7% of the information from the original data. Subsequently, clustering with K-Means was performed using these eight new variables, resulting in the formation of two clusters.
