GeoShapley: A Game Theory Approach to Measuring Spatial Effects in Machine Learning Models

Abstract

This paper introduces GeoShapley, a game theory approach to measuring spatial effects in machine learning models. GeoShapley extends the Nobel Prize-winning Shapley value framework in game theory by conceptualizing location as a player in a model prediction game, which enables the quantification of the importance of location and the synergies between location and other features in a model. GeoShapley is a model-agnostic approach and can be applied to statistical or black-box machine learning models in various structures. The interpretation of GeoShapley is directly linked with spatially varying coefficient models for explaining spatial effects and additive models for explaining non-spatial effects. Using simulated data, GeoShapley values are validated against known data-generating processes and are used for cross-comparison of seven statistical and machine learning models. An empirical example of house price modeling is used to illustrate GeoShapley's utility and interpretation with real world data. The method is available as an open-source Python package named geoshapley.

arXiv comment30 pages, 10 figures, 6 tables

References in corpus (8)

• Scikit-learn: Machine Learning in Python
• Interpretable machine learning: definitions, methods, and applications
• European Union regulations on algorithmic decision-making and a "right to explanation"
• Advances in Hyperspectral Image Classification: Earth monitoring with statistical learning methods
• ETA Prediction with Graph Neural Networks in Google Maps
• A Review of Location Encoding for GeoAI: Methods and Applications
• Machine Learning Estimation of Heterogeneous Treatment Effects with Instruments
• Causal Shapley Values: Exploiting Causal Knowledge to Explain Individual Predictions of Complex Models