Developing Novel Features Employed by Regression Methods to Solve Air Traffic Delay Prediction Problems
Engineering, Technology & Applied Science Research · 04.04.2026 · 16 (2) · 34389-34395
DOI: 10.48084/etasr.17672 ↗Kurz gefasst
Fachartikel in der Zeitschrift Engineering, Technology & Applied Science Research, online erschienen am 04.04.2026. Verfasst von Loc Tran und 5 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Flight delays are influenced not only by flight-specific factors but also by network effects, where disruptions at central airports propagate through downstream connections. This study proposes a lightweight, model-agnostic framework that injects network structure into standard flight-level regression for arrival delay prediction. Using Bureau of Transportation Statistics (BTS) on-time performance data (523,435 completed flights from 2025-01-01 to 2025-01-31), the study builds (i) a directed airport-route graph weighted by observed origin–destination counts and (ii) a rotation hypergraph in which each (tail number, day) forms a hyperedge connecting all airports visited, weighted by the number of legs. From these structures, four airport-level ranking signals are derived: Graph PageRank (GPR), GCN-Style Fixed-Point Ranking (GGCN), Hypergraph PageRank (HPR), and Hypergraph Fixed-Point Ranking (HGCN). Origin and destination ranking features are attached to each flight, and Linear Regression (LR), K-Nearest Neighbors (KNN) regression, and Multi-Layer Perceptron (MLP) regression models are evaluated under a time-aware split. Network-enriched features yield modest but consistent gains, with the best performance achieved by MLP using all ranking features.
Zitieren
APA
Tran, L., Nguyen, L. M. T., Nguyen, H. H., Tran, T. P., Phan, K. A., & Tran, L. (2026). Developing Novel Features Employed by Regression Methods to Solve Air Traffic Delay Prediction Problems. Engineering, Technology & Applied Science Research, 16(2), 34389-34395. https://doi.org/10.48084/etasr.17672
Vancouver
Tran L, Nguyen LMT, Nguyen HH, Tran TP, Phan KA, Tran L. Developing Novel Features Employed by Regression Methods to Solve Air Traffic Delay Prediction Problems. Engineering, Technology & Applied Science Research. 2026;16(2):34389-34395. doi:10.48084/etasr.17672
BibTeX
@article{Tran2026,
author = {Tran, Loc and Nguyen, Luong Minh Tri and Nguyen, Huu Hiep and Tran, Tan Phuc and Phan, Kim Anh and Tran, Linh},
title = {Developing Novel Features Employed by Regression Methods to Solve Air Traffic Delay Prediction Problems},
journal = {Engineering, Technology & Applied Science Research},
year = {2026},
volume = {16},
number = {2},
pages = {34389--34395},
doi = {10.48084/etasr.17672}
}
EUVIMED