An automatic ensemble machine learning for wheat yield prediction in Africa.

Wheat is an essential crop for food security in North Africa. However, it's productivity is limited by several factors, among them climate change effects. Predicting wheat yield on a large scale is thus important for ensuring food security, as it allows farmers and policymakers to make informed decisions regarding agricultural production and marketing. Machine learning (ML) algorithms have been used in previous studies to predict wheat yield in Africa, but there is still a need for improvement in terms of accuracy and usability. The advantage of using several ML algorithms is that it allows for comparing and selecting the best-performing model. The objective of this study is to develop an accessible and user-friendly web-based application that predicts wheat yield using an ensemble learning model that integrates four feature scaling algorithms (Min-Max, Z- score, MaxAbsScaler, and Robust scaling) and seven ML techniques (LASSO, Extreme Gradient Boosting [XGBoost], Random Forest [RF], Linear Regression [LR], Ridge, Gradient Boosting Regression [GBR], Support Vector Regression [SVR]), based on meteorological data (rainfall and temperature), and agriculture and soil properties (pesticides, fertilizers, and irrigation). Findings show that the GBR algorithm with MaxAbsScaler feature scaling is the best with an R 2 of 97%, demonstrating the developed model's effectiveness. Farmers and farm managers could use the suggested model for better decisions, contributing to food productivity and security in North Africa. [ABSTRACT FROM AUTHOR]