Prediction Model of the Burkholderia glumae Pest in Rice Crops Using Machine Learning and Spatial Interpolation

Abstract

In recent years, agriculture, especially rice crops, has been affected by constant climatic changes. Climatic changes cause some pathogens to be favored by altering their products and generating economic losses. The study aimed to develop a machine learning model to predict the appearance of the Burkholderia glumae pest in rice crops in the San Martin region, Peru. The IDW spatial interpolation technique was used to obtain temperature and precipitation data in exploring the data. The study applied a series of supervised algorithms. Among these, the random forest classifier (RFC) was the one that obtained the highest value with an accuracy of 88%. In addition, an application was created where the prediction of the Burkholderia glumae plague in a specific area of the region is displayed.