S-SIRUS: an explainability algorithm for spatial regression Random Forest

Random Forest (RF) is a widely used machine learning algorithm known for its flexibility, user-friendliness, and high predictive performance across various domains. However, it is non-interpretable. This can limit its usefulness in applied sciences, where understanding the relationships between predictors and response variable is crucial from a decision-making perspective. In the literature, several methods have been proposed to explain RF, but none of them addresses the challenge of explaining RF in the context of spatially dependent data. Therefore, this work aims to explain regression RF in the case of spatially dependent data by extracting a compact and simple list of rules from an RF that explicitly takes into account the spatial correlation, i.e. RF-GLS. In this respect, we propose S-SIRUS, a spatial extension of SIRUS, the latter being a well-established regression rule algorithm able to extract a stable and short list of rules from the classical regression RF algorithm. To our knowledge, S-SIRUS is the only explainability tool proposed to open an RF-GLS, which, in turn, is the only random forest algorithm in the literature that accounts for spatial correlation internally in the algorithm. A simulation study was conducted to evaluate the explainability capability of the proposed S-SIRUS, by considering different levels of spatial dependence among the data. The results suggest that S-SIRUS exhibits a higher test predictive accuracy than SIRUS when spatial correlation is present. We encourage the use of SIRUS in the absence of spatial correlation and recommend adopting S-SIRUS when such correlation is present.