A Fuzzy-Based Support Vector Regression Framework for Crop Yield Prediction

This paper proposes a fuzzy-based support vector regression framework for crop yield prediction. To achieve its objectives, interval type-2 fuzzy logic (IT2-FL), principal component analysis (PCA) and support vector regression (SVR) algorithms are employed. The IT2-FL algorithm is used to predict the missing predictor parameter values in the original crop yield prediction dataset while PCA performs feature selection and dimensionality reduction, thereby eliminating redundant information (features) from the transformed dataset by IT2-FL. SVR, machine learning algorithm, is used for model training and testing of the reduced dataset. The performance of the SVR model is evaluated using mean square error (MSE) and root mean square error (RMSE) metrics. Results show that SVR is more accurate at predicting crop yield with an error of 0.002071 for MSE and 0.045513 for RMSE, providing robustness to outliers and minimizing generalization errors with accuracy of 99% for MSE and 95% for RMSE. This indicates that farmers can rely on predicted outcomes from the proposed framework to adopt practices that maximize crop production, improve crop yield, and sustain food sufficiency to the teeming population.