An international co-authored paper has been accepted and published in the international journal “Transportmetrica A: Transport Science”
An international co-authored paper by Rsch. Assoc. Hiroyuki Hasada (first author) and Assoc. Prof. Yudai Honma (corresponding author) from our laboratory has been accepted for publication in Transportmetrica A: Transport Science, an international journal in the field of transportation science, and was published online on March 16, 2026. The title of the paper is “Unveiling implicit path outlier deviations via foundational linear inverse optimization in mobility networks.”
In this study, we focused on the irregularities and suboptimal routes present in observed route data. By formulating the inverse shortest path problem as a linear programming problem, we proposed a new framework that detects structurally anomalous routes while estimating latent link costs. Results from validation on both synthetic and real-world networks demonstrated higher recall rates than existing methods, particularly in the detection of spatially adjacent anomalous routes. This research provides a foundation for the analysis of increasingly sophisticated mobility networks, paving the way for real-time applications and AI-based outlier detection.
