Weather Based Prediction Models for Disease and Pest Using Machine Learning: A Review

Abstract

Critical review of weather based prediction models of disease and pest attack on crops using machine learning (ML) algorithms are performed in the study. Since suitable weather conditions are the accelerators for the growth and spreading of disease or pest, the prediction models based on weather condition achieves high degree of accuracy. Due to the advancement of technology ML algorithms remarks successful application in prediction of diseases and pest on crops. The scope of the review work lies in the fact that the accurate forewarning system helps for the timely application of pest and disease management techniques which have greater significance in controlling and solving the damages due to diseases or pest infestation in plants. Stages in prediction models are analysed and the applied techniques are compared in detail in this review. Consequently, importance of weather parameters in perdition and, performance metrics used for evaluating the prediction models are compared and presented. The review presents the detailed discussion on machine learning algorithms used in the prediction models. The review reveals that new models with high degree of accuracy need to be developed for the prediction of diseases or pest outbreak of various crops.