Comparative analysis of machine learning algorithms for predicting sugarcane yield: insights from recent literature

MIT, Serafin C. Palmares and DIT, Patrick D. Cerna (2025) Comparative analysis of machine learning algorithms for predicting sugarcane yield: insights from recent literature. International Journal of Science and Research Archive, 14 (3). pp. 264-269. ISSN 2582-8185

Predicting sugarcane yield is critical in precision agriculture, particularly in the Philippines, where sugarcane is a cornerstone of the agricultural economy, contributing significantly to sugar production and biofuel generation. This paper provides a comprehensive comparative analysis of various machine learning (ML) algorithms used for predicting sugarcane yield, drawing insights from recent Philippine-based literature from 2021 to 2024. The study evaluates the performance of regression-based and ensemble learning models, including Linear Regression (LR), Support Vector Regression (SVR), Random Forest (RF), and Gradient Boosting Machines (GBM), highlighting their effectiveness, challenges, and future research directions. Performance metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), R-squared (R²), and computational complexity are analyzed to determine the most effective techniques for improving sugarcane yield estimation and farm productivity. The findings aim to assist researchers and agricultural stakeholders select optimal predictive models tailored to the Philippine context, addressing challenges such as data accessibility and computational resource limitations.