Abstract

Rice production forecasting is critical for food security planning in the Philippines, yet existing estimation methods in provinces like Pampanga remain reliant on historical averages and manual assessments. This study presents A-Rice, a sequential hybrid machine learning model that integrates Random Forest (RF) and Gaussian Process Regression (GPR) for municipal-level rice production prediction across 20 municipalities over a 25-year period (2000–2025). The model was developed using historical agricultural data from the Philippine Statistics Authority and the Office of the Provincial Agriculturist, combined with six environmental variables from NASA POWER. Bayesian Small Area Estimation was applied to reconstruct missing municipal-level records for 2000–2016. In the hybrid framework, RF first generates baseline predictions and identifies key predictive features, after which GPR refines these predictions by modeling residual errors and quantifying uncertainty through calibrated confidence intervals. Evaluated against standalone RF and GPR models, the hybrid RF→GPR model achieved R² values exceeding 0.99 across all data splits, with an 80% reduction in RMSE on the test set (from 5,555 to 1,130 metric tons). Feature importance analysis revealed that municipality, cropping season, and relative humidity were the most influential predictors. The model was deployed as a web-based decision-support application providing production forecasts with confidence intervals, environmental diagnostics, and historical benchmarking. Results demonstrate the viability of sequential hybrid ML approaches for precision agriculture and data-driven agricultural planning in developing regions.

References

• Philippine Statistics Authority. 2023. OpenSTAT: Philippine Statistical Database. https://openstat.psa.gov.ph/.
• National Economic and Development Authority Regional Office III. 2023. Central Luzon Regional Development Plan 2023–2028. City of San Fernando, Pampanga.
• Wolanin, A., Mateo-García, G., Camps-Valls, G., et al. 2020. Estimating and Understanding Crop Yields with Explainable Deep Learning in the Indian Wheat Belt. Environmental Research Letters, 15(024019), 1-12.
• Basha, S. M., Rajput, D., et al. 2020. Crop Yield Prediction Using Random Forest. Scalable Computing: Practice and Experience, 21(4), 593–602.
• Ghosh, S. S., Dey, S., et al. 2022. Gaussian Process Regression Model for Crop Biophysical Parameter Retrieval from Multi-Polarized C-Band SAR Data. Remote Sensing, 14(4), 934-960.
• Martínez-Ferrer, L., Piles, M., and Camps-Valls, G. 2021. Crop Yield Estimation and Interpretability with Gaussian Processes. IEEE Geoscience and Remote Sensing Letters, 18(12), 2043-2047.
• Lagrazon, P. G. and Tan, J. 2024. Predicting Crop Yield in Quezon Province, Philippines Using Gaussian Process Regression. 2023 International Conference on Modeling & E-Information Research, 6-12.
• Hariyani, G., Karad, V., et al. 2024. Analysis on Crop Yield Prediction using Various Ensemble Methods. 8th International Conference on Computing, Communication, Control and Automation, 1-6.
• Jabed, Md. A. and Azmi Murad, M. A. 2024. Crop Yield Prediction in Agriculture: A Comprehensive Review of Machine Learning and Deep Learning Approaches. Heliyon, 10(24), e40836.
• Rashid, M., Bari, B. S., et al. 2021. A Comprehensive Review of Crop Yield Prediction Using Machine Learning Approaches. IEEE Access, 9, 63406–63439.
• Shawon, S., Ema, F., et al. 2023. Crop Yield Prediction: Robust Machine Learning Approaches for Precision Agriculture. 26th International Conference on Computer and Information Technology, 1-6.
• Zhou, W., et al. 2022. Random Forest-based Rice Yield Prediction in Mountainous Regions. Agricultural Systems.
• Parreño, S. J. E. and Anter, M. C. J. 2024. New Approach for Forecasting Rice and Corn Production in the Philippines Through Machine Learning Models. Multidisciplinary Science Journal, 6(9), 2024168.
• Choudhary, K., et al. 2022. Random Forest for Rice Yield Prediction Using Sentinel-2 Imagery and Environmental Variables.
• Köse, U., et al. 2023. Random Forest Outperformance in Rice Yield Prediction Using Climatic and Historical Data.
• Liu, F., Su, L., et al. 2024. Prediction Models of Growth Characteristics and Yield for Chinese Winter Wheat Based on Machine Learning. Agronomy, 14, 839–856.
• Ekanayake, P. and Wickramasinghe, L. 2022. GPR-based Paddy Yield Prediction Using Climate Variables in Sri Lanka.
• Hassan, M. A., et al. 2025. GPR Effectiveness in Mechanized Rice Farming Yield Forecasting.
• Ghosh, M. and Rao, J. N. K. 1994. Small Area Estimation: An Appraisal. Statistical Science, 9(1), 55–93.
• Keen, P. G. W. and Scott Morton, M. S. 1978. Decision Support Systems: An Organizational Perspective. Addison-Wesley.
• Power, D. J. 2002. Decision Support Systems: Concepts and Resources for Managers. Quorum Books.
• Sprague, R. H. and Carlson, E. D. 1982. Building Effective Decision Support Systems. Prentice Hall.
• Mabunga, Z. and Dela Cruz, J. 2022. An Optimized Soil Moisture Prediction Model for Smart Agriculture Using Gaussian Process Regression. IEEE 18th International Colloquium on Signal Processing & Applications, 243-247.
• Alebele, Y., Wang, W., et al. 2021. Estimation of Crop Yield From Combined Optical and SAR Imagery Using Gaussian Kernel Regression. IEEE JSTARS, 14, 10520–10534.