Surrogate Graph Partitioning for Spatial Prediction
This work addresses the need for interpretable spatial prediction in practical applications, though it is incremental as it builds on existing surrogate model methods.
The paper tackles the problem of making spatial prediction models interpretable for industries requiring transparency by proposing a graph partitioning approach to create spatial segments that minimize within-segment prediction variances, with experiments showing computational efficiency in identifying these segments.
Spatial prediction refers to the estimation of unobserved values from spatially distributed observations. Although recent advances have improved the capacity to model diverse observation types, adoption in practice remains limited in industries that demand interpretability. To mitigate this gap, surrogate models that explain black-box predictors provide a promising path toward interpretable decision making. In this study, we propose a graph partitioning problem to construct spatial segments that minimize the sum of within-segment variances of individual predictions. The assignment of data points to segments can be formulated as a mixed-integer quadratic programming problem. While this formulation potentially enables the identification of exact segments, its computational complexity becomes prohibitive as the number of data points increases. Motivated by this challenge, we develop an approximation scheme that leverages the structural properties of graph partitioning. Experimental results demonstrate the computational efficiency of this approximation in identifying spatial segments.