Pengklasteran dan Penerapan Rantai Markov untuk Prediksi Produksi padi dan Lahan Panen di Kalimantan Barat

Aisyah Ulfah (Departemen Statistika, Fakultas Sains dan Matematika, Undip)
Sudarno sudarno (Departemen Statistika, Fakultas Sains dan Matematika, Undip)
Rahmila Dapa (Departemen Statistika, Fakultas Sains dan Matematika, Undip)

Jurnal Gaussian
Vol 14, No 2 (2025): Jurnal Gaussian

The population growth rate in Indonesia is very high. This causes the need for rice to also increase. The increase in population growth is not matched by the growth of harvested land. This causes the area of harvested land to decrease because a lot of land is converted into settlements. Rice production and harvest land in West Kalimantan has decreased in the period 2019 to 2021. The purpose of this study is to classify and predict rice production and harvest land in West Kalimantan. In this research, the clustering method used is Cluster Time Series with Average Linkage method. Average Linkage is one of the hierarchical groupings based on the average distance between objects. The Silhouette Coefficient value is used to determine the optimal number of clusters. This research also uses the Markov Chain method to predict rice production and harvestable land. Markov chain is a stochastic process that explains future events only depend on today's events and do not depend on past circumstances. The results of this study obtained two clusters and many districts/cities are in clusters that have low rice production and harvest land. The prediction results of rice production and harvested land in Kalimantan have the greatest chance of experiencing a decline.