EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Unveil Fundamental Graph Properties for Neural Architecture Search

Zhenhan Huang, Tejaswini Pedapati, Pin-Yu Chen, Chunheng Jiang, Jianxi Gao iD

Advanced Science · 2026-05 · 13 (25)

DOI: 10.1002/advs.202516574 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advanced Science, online erschienen am 23.02.2026. Verfasst von Zhenhan Huang und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, Rensselaer Polytechnic Institute, Troy, USA. Die Arbeit nennt 86 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Science Foundation, IBM Research.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Deep learning profoundly impacts various areas, such as face recognition and language translation. Owing to the increasingly high computational costs of training neural architectures, it is intractable to manually examine the performance of various neural architectures, promoting the area of Neural architecture search (NAS) that enables artificial intelligence (AI) automation. Despite the advance of NAS in automatically discovering the optimal neural networks, the fundamental understanding of the structure of neural architectures is limited. To fill the gap, we propose a method dubbed NASGraph that converts neural architectures to graphs whose graph properties determine their performances, enabling us to search for or design strong architectures. On various standard NAS benchmarks, NASGraph outperforms baseline NAS methods and requires remarkably fewer computation resources. The numerical results illuminate the relationship between neural architecture and its performance, complementing other approaches. Therefore, combining our approach with these approaches leads to performance improvement, but not combining the others. Our study offers a new perspective on network science for AI, potentially advancing various aspects of machine learning and uncovering the black box of convolutional neural networks.

Schlagwörter

Stichwörter der Autor:innen

AI automationdeep learningnetwork propertyneural architecture search

Autor:innen und Institutionen

  • Department of Computer Science, Rensselaer Polytechnic Institute, Troy, USA.
  • IBM Research, Yorktown Heights, USA.

ORCID-Profile: Jianxi Gao

Förderung

  • National Science Foundation
  • IBM Research

Zitieren

APA

Huang, Z., Pedapati, T., Chen, P. Y., Jiang, C., & Gao, J. (2026). Unveil Fundamental Graph Properties for Neural Architecture Search. Advanced Science, 13(25). https://doi.org/10.1002/advs.202516574

Vancouver

Huang Z, Pedapati T, Chen PY, Jiang C, Gao J. Unveil Fundamental Graph Properties for Neural Architecture Search. Advanced Science. 2026;13(25). doi:10.1002/advs.202516574

BibTeX
@article{Huang2026,
  author = {Huang, Zhenhan and Pedapati, Tejaswini and Chen, Pin-Yu and Jiang, Chunheng and Gao, Jianxi},
  title = {Unveil Fundamental Graph Properties for Neural Architecture Search},
  journal = {Advanced Science},
  year = {2026},
  volume = {13},
  number = {25},
  doi = {10.1002/advs.202516574}
}
Kostenlos starten →