Prediction of Crop Disease Using Rank Regressive Learning and Proaftn Fuzzy Classification Models

Abstract

Today, Machine Learning (ML) is driving the big variation covering significant industries. Agriculture is one industry where ML researchers are operating with farmers to assist farmers put together a better and more significant utilization of the dwindling resources owing to the metropolitan and big data concept. Nevertheless, plant disease, specifically crop plants, is an extensive menace as far as global food security is concerned. Diseases prevalent in plants at the first hand influence the fruit or grain quality, therefore resulting in the lessening of agricultural fertility. The conventional method of crop disease identification is performed via visual investigation. However, this type of process is said to be extremely incompetent and hence susceptible to the error with the inception of big data. Over the past few years, several works on ML techniques for crop disease prediction have been proposed. To overcome this proposed work, a novelty method called Rank Regressive Learning and Proaftn Fuzzy Classification - crop disease prediction with big data for soybean is proposed. The RRL-PFC crop disease prediction method is split into two sections. They are feature selection and crop disease prediction. Initially, raw data is obtained from the Soybean dataset (i.e., one of the crops utilized from crop yield), that are produced to input vector matrix. To choose the relevant features that are designed for categorizing crop disease, the Rank Regressive learning-based Feature Selection is applied than vector matrix as input. Finally, with the computationally efficient selected features, categorization of crop disease is made by means of Proaftn Fuzzy Classification-based Crop Disease Prediction in an accurate manner. The experimental evaluation of the proposed RRL-PFC method with respect to improved accuracy and lesser time, overhead, and error rate than the conventional methods.