Prediction of Pest Infestation in Tea Leaves Using Machine Learning Models

Climate change and resource depletion provide several possible challenges to the sustainable production of agricultural products and their management. Crop production systems are threatened by weeds, insects, and pests, significantly reducing yields. The agricultural industry is experiencing a noticeable impact of climate change on the development and yield of plants, including tea. This study states the effects of key meteorological parameters on tea plant diseases and pests. Pest infestation is a major problem in this region, and it can lead to significant losses for farmers. Machine learning regression models, namely random forest (RF), support vector machine (SVM), and K-nearest neighbor (KNN), were developed for the prediction of pest infestation percentage in tea plantations using meteorological parameters. This model can then help farmers take preventive measures to reduce pest infestation. The Pearson correlation method was used for selecting the weather parameters that majorly affect the infestation percentage (IP). The RF model gave the best results for predicting IP with the R2 values for training and testing were 0.9240 and 0.5790, and MAE values of 1.5441% and 2.5842%, respectively. The developed ML-based pest IP prediction model can be used for the different tea plantation regions to predict IP in advance.