Gaussian Mixture Model for Immigration Residence Permit in Kediri Immigration Office


  • Priati Assiroj
  • Besse Hartati
  • Isidorus Anung
  • Nurul Maharani
  • Rasona Sunara
  • Galuh Boy
  • Masdar Bakhiar
  • Atsil Syah
  • Tiara Okta


Data mining, Residency permit data, Cluster, Gaussian Mixture Model (GMM)


Data mining involves the utilization of pattern recognition techniques, mathematical concepts, and machine learning algorithms to uncover valuable trends and relationships within datasets. Cluster analysis is a fundamental method in data mining, aiding in identifying clusters with similarities. The use of clustering algorithms has garnered attention across various data analysis domains, and model-based clustering techniques have gained traction in managing complex data. The proposed model-based clustering approach combines various non-continuous variables using the Generalized Linear Latent Variable Model (GLLVM) and Gaussian Mixture Model (GMM) to analyze mixed data. This approach has been extended to consider latent variables with non-Gaussian distributions, resulting in the natural formation of clusters. This research highlights the importance of utilizing data mining techniques to reveal insights and patterns within complex datasets, contributing to enhanced data management and informed decision-making. The dataset used is residency permit data, resulting in 2 clusters that are in line with the DBI, providing information that Cluster 1 is predominantly composed of foreign nationals (WNAs) sponsored by corporations for residency or visits to the Kediri area and its surroundings. On the other hand, Cluster 2 is characterized by WNAs sponsored by individuals or without sponsors, with their destinations lying outside the Kediri district