Fertilizer Classification for Identifying Effectively Commodity to The Crop Land Using Naïve Bayes and Decision Tree Classifier

Agriculture has a crucial role in meeting the food needs of the community and other farmer groups. One of the efforts to empower farmers is through farmer groups that aim to increase productivity, income, and welfare among farmers. Farmer groups must develop communities as a condition for obtaining subsidized fertilizer from the government. This study uses machine learning classifier methods, especially ID3 and Naïve Bayes algorithms, to identify farmer group communities’ data and predict subsidized fertilizer orders for subsectors and commodities. The evaluation show that both classification algorithms have a high accuracy rate, reaching about 99%. The ID3 algorithm provides high precision for the Urea fertilizer class, while Naïve Bayes has a lower precision for the nitrogen, phosphorus, and potassium (NPK) class. However, Naïve Bayes showed a higher recall value for the NPK class. This algorithm provides valuable insights and information to more effectively manage farm resources. The results of this research are expected to increase agricultural output, improve the welfare of farmers, and support community food security. In conclusion, this study successfully used machine learning to predict subsidized fertilizer orders for subsectors and commodities. Both algorithms give accurate results, and the use of any of them depends on individual preferences and needs.