Implementasi SOM Dalam Clustering Hasil Ikan Laut Kabupaten Pekalongan

Abstract

Data from sea fish in Pekalongan Regency can be processed, one of which is clustered. Clusters are grouping data based on the same criteria. The purpose of doing clustering is to be able to help in sorting and dividing a situation based on the same criteria. Clustering of marine fish products in Pekalongan Regency will be grouped into three groups, namely: a small group of marine fish products, a medium group of marine fish products, and a large group of marine fish products. The clustering process uses the SOM algorithm, and the data is taken from the website data.go.id/dataset. Data is processed in order to show which fish yields are small, medium and large. The processing process uses variable types of fish, years and results of sea fish that are stored in Excel files and then processed using Matlab. The results show that there are fish species that are classified as low and moderate clusters, namely shrimp, squid, serimping, grouper, turmeric, and ray species. The types of fish that enter the cluster and many are Tigawaja. The types of fish that enter the medium cluster are Beloso, Pihi, Pepetek, and those who enter the low cluster are 18 fish species, while those who enter the low, medium and many clusters are Petek.