Machine Learning based Classification of Diseased Mango Leaves

Abstract

The preponderance of population depends on agriculture to produce crops which would be their primary subsistence for their livelihood. So, agriculture is considered the backbone of any nation. Mango (Mangifera indica Linn), belonging to a family Anacardiaceous, is a conspicuous fruit that captivates all ages because of its meticulous taste, delicious flavor, ampleness variety, and highly lustiness. Mangoes are generally rich in minerals, vitamins, fibers, and negotiable fat. Mango plants are exposed to many micro-organisms. If these are not detected and treated in the initial developing stages, it would affect peculiar parts of the mango plant and result in loss of overall productivity. Several factors like biotic and abiotic always ensue in the decrease in the overall productivity of mango plants. Self-regulated Detection of mango plant disease is imperative, and it must be detected at the preliminary stages of the growing period of the mango plant. This paper discusses the existing methodology to classify diseases in mango plant leaves by implementing ensemble technique (Stack) which includes algorithms like Decision Tree (DT), Support vector machine (SVM), Neural Network (NN), and Logistic Regression (LR). The developmental results validate that the disease classification methodology can successfully classify a higher percentage in predicting whether mango plant leaf is healthy or diseased.