Mapping of rice commodities margin trading in Indonesia using unsupervised learning algorithm

• Main Authors: Badaruddin, Muliati, Buna, Abdul Malik I., Puspa, Misrawati Aprilyana, Santawali, S., Anwar, Rapika, Hutahaean, Jeperson
• Format: Conference Proceeding
• Language: English
• Subjects: Algorithms; Clustering; Commodities; Machine learning; Unsupervised learning
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0202115

Rice is a strategic product that is vital to the food security of Indonesia. This is because rice is the primary food consumed daily by the majority of Indonesians. The fact that rice consumption accounts for 20.03 percent and 24.06 percent, respectively, of the urban and rural poverty lines demonstrates how dependent the Indonesian population is on this staple. This study aims to determine how trade and transportation margins for rice commodities in Indonesia vary by province. Trade and Transportation Margin is the difference between total sales value and total purchases value. Clustering Margin of Rice Commodity Trade and Transport is calculated using official data from the Central Bureau of Statistics for 2018-2020. The unsupervised learning technique with K-means as the algorithm is utilized to classify objects in each province. The findings of the study indicated that the K-means approach may be utilized to map the Trade and Transport of Rice Commodities by Province in Indonesia. The presence of nine provinces in the top cluster indicates that this province would generate a profit between 2018 and 2020. The value of the three suggested clusters was -1.026.1 We may conclude based on the DBI value that cluster 7 is the best in this investigation.