Explainable Demand Prediction for Ride-Hailing Services
Elif Arslan (TU Delft - Civil Engineering & Geosciences)
Emir Demirović (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Marco Rinaldi (TU Delft - Civil Engineering & Geosciences)
Oded Cats (TU Delft - Civil Engineering & Geosciences)
Serge Hoogendoorn (TU Delft - Civil Engineering & Geosciences)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The emergence of ride-hailing services has introduced computational challenges in optimizing vehicle routes. One of these challenges is solving the routing problem for long-term periods, as it requires accurate and explainable demand predictions. This is because explainability influences solution trustworthiness while accuracy determines final optimization outcomes. To address this, we employed Multi-Objective Optimal Regression Tree (MOO-RT) with accuracy and explainability objectives. Comparative analysis against other models, utilizing accuracy (MSE) and explainability (global Shapley Value) metrics, demonstrated MOO-RT’s impact on enhancing explainability whereas Random Forest emerged as the top performer in accuracy. Moreover, using Shapley values our study identified model-independent significant features for passenger prediction.