Spatial boosting method combines maps and data for faster insurance risk prediction
SAMBoost: Spatially Aware Mesh Boosting
Machine Learning
Summary
Estimating how risks vary across locations is important but challenging when also accounting for other factors like weather or property details. The authors created SAMBoost, a new method that mixes tree-based models with spatial maps to predict risks more accurately and quickly. Unlike older methods that trade off speed or detail, SAMBoost captures both the geographic patterns and the effects of other factors at the same time. It was tested on big insurance data and found to be faster and more precise than existing tools.
What this means in practice
- •For insurance data teams: Train comprehensive risk models that reveal interpretable geographic risk patterns alongside detailed customer factors at scale.
- •For environmental risk analysts: Generate accurate spatial risk surfaces from large environmental datasets while capturing complex interactions with other measured variables.
Authors
Mulah Moriah, Franck Vermet, Pierre Ailliot, Juliette Legrand
Abstract
We propose SAMBoost (Spatially Aware Mesh Boosting), a gradient boosting method that combines tree-based covariate modelling with a stochastic partial differential equation (SPDE) latent spatial field through a single penalised objective. At each boosting iteration, the predictor is decomposed into separate covariate and spatial components. The method is motivated by applications in which the spatial effect is not a nuisance to be removed but a quantity of interest that must be recovered alongside the covariates. In insurance, for example, practitioners need both flexible nonlinear covariate effects and an interpretable geographic risk surface that can be reported on its own. Existing approaches often involve trade-offs, offering flexible covariate modelling but no explicit spatial field, rigorous spatial inference with a fixed additive covariate structure, or a combination of both at computational costs that limit practicality for large datasets. SAMBoost addresses these goals within the boosting loop, and the underlying mesh produces a sparse precision structure whose cost is largely independent of sample size. On simulations and on an open benchmark, SAMBoost matches or improves on the predictive accuracy of joint Gaussian process and tree boosting approaches. It is also faster and recovers the spatial field and covariate effects more accurately. When covariate effects include interactions, it outperforms additive Bayesian spatial models. Applied to a flood insurance portfolio with 4.9 million records, SAMBoost trains on the full training set without subsampling and runs about sixteen times faster than a Bayesian reference, with a consistent spatial risk surface. A public Python implementation is available.