A brief study on the prediction of crop disease using machine learning approaches

Abstract

The intention of this entire survey is to evaluate the importance and impact of the articles which have been posted with the identify device gaining knowledge of-primarily based totally early detection of crop disorder or prediction of fungal illnesses on vegetation with the help of device gaining knowledge of and facts mining techniques at some stage in the duration 2016-2020. It likewise uncovers that the territory of plant disorder has gotten elevated and hobby with the aid of using researchers, studies investment institutions, and experts. The electronically available peer-review journal papers from Google Scholar, Web of Sciences, and papers available at Mendeley desktop application databases were reviewed. The following parameters were considered while reviewing the papers. 1. Which machine learning or data mining algorithmic approach was used? 2. Which performance metrics were used? 3. Which plant diseases data set was used? 4. How was the performance analysis carried out? 5. Whether the results were compared with some other techniques? The computer algorithms-based articles deal with the early detection of plant disease and were published between 2016 and 2020 were reviewed. From the top-refered to explore distributions relating to AI based expectation of plant infection, it is seen that mixture models were broadly used over a singular order model. A broadly utilized relapse model with SVM, variations of choice trees, and Naive Bayes models are having the best exhibition for early expectation of yield infections.