Gaussian process regression as an alternative to kriging and SVM for spatial yield prediction
Keywords:
modeling, geostatistics, machine learning, interpolation, spatial analysisAbstract
Detecting spatial yield variability is essential for precision agriculture because it minimizes environmental impact and enhances economic returns. This study proposes Gaussian Process Regression (GPR) as a predictive model for yield estimation, particularly in cultivated areas where the highest yields appear in the central region of the field, while the edges exhibit lower productivity. The study was conducted in Patos de Minas, Brazil, using 795 georeferenced soybean yield samples over 3.7 hectares. The analysis evaluates GPR across different sample sizes and compares it with Ordinary Kriging (OK) and Support Vector Machine (SVM). The results indicate that GPR and OK perform similarly under high sampling densities, but GPR achieves higher predictive accuracy under low-sampling conditions. A sample size of at least 60% of the full dataset is necessary to maintain reliable spatial prediction, as
smaller sample sizes lead to greater prediction errors and less defined spatial patterns. SVM, in contrast, produces a smoothing effect across all sampling densities, which reduces its ability to capture local variations. These findings highlight GPR as a robust alternative for yield mapping, particularly in scenarios with limited data availability. From a practical perspective, GPR and OK remain strong candidates for yield prediction, reinforcing the importance of model selection based on data availability and spatial variability.
References
1. Lucas MP, Longman RJ, Giambelluca TW, Frazier AG, McLean J, Cleveland SB, et al. Optimizing automated kriging to improve spatial interpolation of monthly rainfall over complex terrain. J Hydrometeorol. 2022;23(4):561-72. doi:10.1175/ JHM-D-21-0171.1.
2. Sifaou H, Kammoun A, Alouini MS. A precise performance analysis of support vector regression. In: Proceedings of the 38th International Conference on Machine Learning (ICML); 2021. p. 9671-80.
3. Siqueira DS, Barros JGA, Marques JS, Pereira GT. Geostatistical and machine learning models for spatial prediction of soil properties in precision agriculture. Revista Ceres. 2021;68(5):384- 93. doi:10.1590/0034-737X202168050008.
4. Campos-Taberner M, García-Haro FJ, Moreno Á, Gilabert MA, Sánchez-Ruiz S, Martínez B, et al. Mapping leaf area index with a smartphone and Gaussian processes. IEEE Geosci Remote Sens Lett. 2015;12(12):2501-5. doi:10.1109/LGRS.2015.2488682.
5. Martínez-Ferrer L, Piles M, Camps-Valls G. Crop yield estimation and interpretability with Gaussian processes. IEEE Geosci Remote Sens Lett. 2021;18(12):2043-7. doi:10.1109/LGRS.2020.3016140.
6. Alebele Y, Wang W, Yu W, Zhang X, Yao X, Tian Y, et al. Estimation of crop yield from combined optical and SAR imagery using Gaussian kernel regression. IEEE J Sel Top Appl Earth Obs Remote Sens. 2021;14:10520-34. doi:10.1109/JSTARS.2021.3118707.
7. Martins GD, Xavier LCM, Oliveira GP, Gallo MLBT, Abreu Júnior CAM, Vieira BS, et al. Using geospatial information to map yield gain from the use of Azospirillum brasilense in furrow. Agronomy. 2023;13(3):808. doi:10.3390/agronomy13030808.
8. Alvares CA, Stape JL, Sentelhas PC, Gonçalves JLM, Sparovek G. Köppen’s climate classification map for Brazil. Meteorol Z. 2013;22(6):711-28. doi:10.1127/0941-2948/2013/0507.
9. Climate-Data.org. Climate of Patos de Minas. 2024 [cited 2025 Sep 8]. Available from: https://en.climate-data.org/south-america/ brazil/minas-gerais/patos-de-minas-2893/
10. He H, Zheng T, Zhao J, Yuan X, Sun E, Li H, et al. Improved Gaussian regression model for retrieving ground methane levels by considering vertical profile features. Front Earth Sci. 2024;12:1352498. doi:10.3389/feart.2024.1352498.
11. Rasmussen CE, Williams CKI. Gaussian processes for machine learning. Cambridge (MA): MIT Press; 2006. 266p.
12. Rodrigues BP, Rofatto VF, Matsuoka MT, Assunção TT. Resampling in neural networks with application to spatial analysis. Geospat Inf Sci. 2022;25(3):413-24. doi:10.1080/10095020.2022. 2040923.
13. Pereira GW, Valente DSM, Queiroz DMD, Coelho ALDF, Costa MM, Grift T. Smart-map: an open-source QGIS plugin for digital mapping using machine learning techniques and ordinary kriging. Agronomy. 2022;12(6):1350. doi:10.3390/agronomy12061350.
14. Li J. A critical review of spatial predictive modeling process in environmental sciences with reproducible examples in R. Appl Sci. 2019;9(10):2048. doi:10.3390/app9102048.
15. Chandra MA, Bedi SS. Survey on SVM and their application in image classification. Int J Inf Technol. 2021;13(5):1-11. doi:10.1007/s41870-017-0080-1.
16. Ribeiro JL, Silva AM. Comparative study of kriging and support vector machines for spatial prediction of soil properties. Comput Electron Agric. 2018;155:183-93.
17. Carvalho E, Assis GA, Martins GD, Marques DJ, Santos EA, Xavier LCM, et al. Intercropped soybean plant population in a coffee plantation and its effects on agronomic parameters and geospatial information. Agronomy. 2024;14(2):343. doi:10.3390/ agronomy14020343.
18. Filippi P, Han SY, Bishop TF. On crop yield modelling, predicting, forecasting and addressing the common issues in published studies. Precis Agric. 2025;26(1):8. doi:10.1007/s11119-024-10212-2.
19. Rofatto VF, Philippe S, Martins GD. Gaussian process regression as an alternative to kriging and SVM for spatial yield prediction. SciELO Preprints. 2025. doi:10.1590/SciELOPreprints.11312.
