A New Approach for Paddy Leaf Blast Disease Prediction Using Logistic Regression

Abstract

Paddy is a major agricultural crop. But the production of paddy is hindered by various kinds of diseases. Some of those diseases are leaf blast disease, brown spot disease, bacterial blight disease, etc. Amidst of all these diseases influencing the paddy production, leaf blast disease had a great influence and it is the most destructive diseases that are effecting on paddy crop. Leaf blast is risen by the fungus Magnaporthe oryzae. It will affect all the above-ground parts of a paddy crop: leaf, collar, node, neck, parts of panicle, and sometimes leaf sheath. Thus, examining and accurate forecasting for the development of blast disease are significant and early forecasting of the disease is very beneficial. Many former blast disease prediction models were only considering the attribute values but not their correlations. In this paper, logistic regression algorithm is applied for forecasting the occurrence of leaf blast disease for Adilabad district of Telangana state in India during 2007–2017 in order to prevent the paddy fields from disease. With the help of the correlation mining and clustering process among the attributes, we are classifying the attribute sets based on their impact on disease occurrence. Finally, the logistic regression algorithm calculates the leaf blast disease occurrence probability.