Comparison of classification algorithms in classifying tomato plant disease data (Lycopersium esculentum Mill.)

Tomato plants are an important crop for the agricultural and trade sectors. Tomato plants are a type of fruit that is nutritious and has many benefits for human health. Diseases in tomato plants are a problem for tomato growth and production. Timely and targeted identification is the main and important thing to ensure that crop failure does not occur which results in losses for farmers. By utilizing data mining technology, tomato disease classification can be carried out quickly and accurately. This research aims to classify tomato plant disease data using Decision Tree, Naïve Bayes, and K-nearest neighbor (KNN) algorithms, then compare the performance of these classification algorithms. Algorithms performance is measured through accuracy, precision, recall, and Area Under Curve (AUC) tests using a confusion matrix. The research results concluded that the best performance for classifying tomato plant disease data was the KNN algorithm, with an accuracy value of 0.820 or 82%, a precision value of 0.828 or 83%, a recall value of 0.820 or 82%, an AUC value of 0.968 or 97%.