Classification of Corn Diseases using Random Forest, Neural Network, and Naive Bayes Methods

Abstract

Corn is one of the staple foods consumed by many people after rice plants, especially in Indonesia. High consumer demand requires corn production in large quantities to meet these needs. However, corn production is not always in large quantities due to several factors, namely diseases in corn plants. Unhealthy corn plants can reduce the amount of production. Healthy and unhealthy corn plants can be identified manually, but this method is not efficient, so in this study, it is proposed to classify corn diseases using the Random Forest, Neural Network, and Nave Bayes methods. The dataset used is a collection of corn leaf images taken from farmers’ fields in the Madura Region with four target classes, namely healthy, gray leaf spot, blight, and common rust. Based on the test results, the classification using the Neural Network method provides a better accuracy value than the other two methods in classifying corn leaf datasets, namely the AUC value reaches 90.09%, classification accuracy is 74.44%, f1 score is 72.01%, precision is 74.14% and recall by 74.43%.