Prediction of Wheat Yield in Baghdad Governorate Using Intelligence Support Vector Machine Algorithm

Abstract

This research aimed to analyze the effect of economic, social, technical, administrative, and climatic factors on the prediction of productivity for wheat during the 2023-2024 growing season for (100) farmers were randomly surveyed in Baghdad Governorate using an intelligent machine learning algorithm (Support Vector Machine SVM), and its performance was evaluated using R² and RMSE criteria. Results showed that the relative importance of the hail and educational level variables (HAIL and EDU) reached 43.20% and 39.70%, respectively. Therefore, we conclude that these two variables accounted for a combined percentage exceeding 80%, thus achieving the first and second ranks and demonstrating statistical significance (Sig.) at a significance level of )P < 0.01), This indicates that they were the most influential factors on the dependent variable (PRODUC) Among the other independent variables. We concluded that the SVM model had acceptable fit of Goodness, reliability, and good prediction accuracy, as the results of the performance evaluation criteria represented by (R², RMSE) reached approximately (0.66, 23.89 kg/dunum), so research hypothesis was proven. The research recommends the necessity of resorting to the use of the SVM machine learning algorithm in predicting the productivity of other crops for different agricultural sites.