# AMBIT: Augmenting Mobility Baselines with Interpretable Trees

arXiv: https://arxiv.org/abs/2512.22466 · HTML: https://mlsanity.com/paper/2512.22466/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 24 references that are in the map.
- First posted on arXiv 27 Dec 2025.

## Builds on (most cited first, 7 of 7)

- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- [Deep Gravity: enhancing mobility flows generation with deep neural networks and geographic information](https://mlsanity.com/paper/2012.00489/) (2020)
- [Predicting origin-destination ride-sourcing demand with a spatio-temporal encoder-decoder residual multi-graph convolutional network](https://mlsanity.com/paper/1910.09103/) (2019)
- [Origin-Destination Network Generation via Gravity-Guided GAN](https://mlsanity.com/paper/2306.03390/) (2023)
- [What's the Price of Monotonicity? A Multi-Dataset Benchmark of Monotone-Constrained Gradient Boosting for Credit PD](https://mlsanity.com/paper/2512.17945/) (2025)
- [Monotone Tree-Based GAMI Models by Adapting XGBoost](https://mlsanity.com/paper/2309.02426/) (2023)
- [DIM: Enforcing Domain-Informed Monotonicity in Deep Neural Networks](https://mlsanity.com/paper/2509.21666/) (2025)

## Abstract (arXiv)

Origin-destination (OD) flow prediction remains a core task in GIS and urban analytics, yet practical deployments face two conflicting needs: high accuracy and clear interpretability. This paper develops AMBIT, a gray-box framework that augments physical mobility baselines with interpretable tree models. We begin with a comprehensive audit of classical spatial interaction models on a year-long, hourly NYC taxi OD dataset. The audit shows that most physical models are fragile at this temporal resolution; PPML gravity is the strongest physical baseline, while constrained variants improve when calibrated on full OD margins but remain notably weaker. We then build residual learners on top of physical baselines using gradient-boosted trees and SHAP analysis, demonstrating that (i) physics-grounded residuals approach the accuracy of a strong tree-based predictor while retaining interpretable structure, and (ii) POI-anchored residuals are consistently competitive and most robust under spatial generalization. We provide a reproducible pipeline, rich diagnostics, and spatial error analysis designed for urban decision-making.