Spatial clustering of villages: a solution for agricultural area management: Case study of Aghmiyun agricultural area, East Azerbaijan Province, Iran

Abstract

All of the world's basic needs are provided by the agricultural sector, and it also provides the raw materials required for industrial processes. Thus, the appropriate administration and development of agriculture are crucial. Like a number of other countries, Iran faces numerous social, economic, environmental, institutional, and human challenges in the development and administration of its agricultural sector. For managing the agricultural sector, Iran has established many different kinds of organizational structures. The Agricultural jihad centers work at the lowest level of the Iranian agriculture sector's governmental administration hierarchy. Each agricultural jihad center covers one or more rural district and a number of villages. The villages covered by each center are divided among the experts who work in the center. How to divide up villages among experts is a major problem facing agricultural jihad centers. Typically, the number of villages under expert administration is not homogeneous, and the allocation of villages among experts is not done appropriately. In this study, village classification has been investigated utilizing four different spatial clustering techniques: KMeans, KMedians, KMedoids, and Spectral Cluster. The findings indicate that while all of these techniques are capable of spatial clustering of villages, KMeans and KMedians are the most effective and useful techniques for clustering.