COMPARISON OF K-MEANS AND GAUSSIAN MIXTURE MODEL IN PROFILING AREAS BY POVERTY INDICATORS

The Covid-19 pandemic has led to income degradation of the Indonesia population which potentially triggers poverty. According Indonesian Central Statistics Agency, Province Java is one areas that most affected by especially on economic aspect. In 2020, percentage poor people increased 0.6% from 2019. If this condition ignored for long term, it will have a negative impact hampering national development. As first step in designing strategy mitigating poverty, necessary carry out an appropriate profiling aspect based poverty indicators. This study compares K-Means Clustering and Gaussian Mixture Model (GMM) providing best data grouping clustering indexes, including: connectivity, Dunn, silhouette. GMM generalization include information about covariance structure as well latent centers. We used indicators Agency Java, such line, population, depth index, severity index. results obtained indicate gives with 3 clusters, number members first, second, third 10, 19, 6 respectively.