Pest Forecaster: A Web‐Based Machine Learning Framework for Climate‐Sensitive Pest Risk Prediction

Manuscript submitted January 30, 2026; accepted March 5, 2026; published April 28. 2026

Abstract—The outbreak of pests is a severe problem in farming that leads to massive loss of crop varieties and poses a threat of food insecurity in the global context. Pest risk should be predicted early so that pests could be intervened in time and crops could be protected in a sustainable manner. The research aims at forecasting the occurrence of pests through machine learning, namely, XGBoost and a Sequential Neural Network (SNN). The models were trained using the historical agricultural data, which comprised of crop type, weather conditions and seasonal factors with the supplementary real-time weather data being incorporated with the OpenWeather API to provide real-time predictions. The preprocessing of data consisted of work with missing values, use of SMOTE to balance the classes (in the case of XGBoost) and scaling of features in the case of the neural network. The measures used to assess model performance were accuracy, precision, recall, F1-score, and confusion matrix. The findings revealed that the XGBoost model performed optimally in prediction, which gave good predictions in terms of feature importance where crop type and seasonality were the most highly important predictors. The Sequential Neural Network also had similar performance and could provide the complex relations between weather variables and pest occurrence. This work offers an addition to precision agriculture and helps to make better decisions related to sustainable pest management as it will include timely and precise predictions without the use of IoT devices or image data.

keywords—pest outbreak, machine learning, XGBoost, sequential neural network

Cite: Paul Olotu, Olutayo Boyinbode, Eyiowuawi Abdulateef, Temitayo Balogun,"Pest Forecaster: A Web‐Based Machine Learning Framework for Climate‐Sensitive Pest Risk Prediction," Journal of Advances in Artificial Intelligence, vol. 4, no. 2, pp. 94-103, 2026. doi: 10.18178/JAAI.2026.4.2.94-103

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).