Strong Optimal Classification Trees
Operations Research · 2025-07 · 73 (4) · 2223-2241
DOI: 10.1287/opre.2021.0034 ↗Kurz gefasst
Fachartikel in der Zeitschrift Operations Research, 2025-07. Verfasst von Sina Aghaei und 2 weiteren Autor:innen. Die Arbeit wurde bisher 36-mal zitiert (Crossref) und nennt 56 Literaturangaben.
Abstract (Originalsprache)
Decision trees are among the most interpretable and popular machine learning models that are used routinely in applications ranging from revenue management to medicine. Traditional heuristic methods, although fast, lack modeling flexibility for incorporating constraints such as fairness and do not guarantee optimality. Recent efforts aim to overcome these limitations using mixed-integer optimization (MIO) for better modeling flexibility and optimality, but speed remains an issue. In “Strong Optimal Classification Trees,” Aghaei, Gómez, and Vayanos use integer optimization and polyhedral theory to create an MIO-based formulation with a strong LO relaxation resulting in a 29% speed-up in training time compared with state-of-the-art MIO-based formulations, as well as up to an 8% improvement in out-of-sample accuracy.
Zitieren
APA
Aghaei, S., Gómez, A., & Vayanos, P. (2025). Strong Optimal Classification Trees. Operations Research, 73(4), 2223-2241. https://doi.org/10.1287/opre.2021.0034
Vancouver
Aghaei S, Gómez A, Vayanos P. Strong Optimal Classification Trees. Operations Research. 2025;73(4):2223-2241. doi:10.1287/opre.2021.0034
BibTeX
@article{Aghaei2025,
author = {Aghaei, Sina and Gómez, Andrés and Vayanos, Phebe},
title = {Strong Optimal Classification Trees},
journal = {Operations Research},
year = {2025},
volume = {73},
number = {4},
pages = {2223--2241},
doi = {10.1287/opre.2021.0034}
}
EUVIMED