EUVIMEDEUVIMED entdecken
Crossref

An Efficient Multi-Objective Evolutionary Zero-Shot Neural Architecture Search Framework for Image Classification

Jianwei Zhang, Lei Zhang, Yan Wang, Junyou Wang, Xin Wei, Wenjie Liu

International Journal of Neural Systems · 2023-05 · 33 (05)

DOI: 10.1142/s0129065723500168 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Neural Systems, online erschienen am 10.02.2023. Verfasst von Jianwei Zhang und 5 weiteren Autor:innen. Beteiligte Einrichtung: College of Computer Science, Sichuan University, Section 4, Southern 1st Ring Road, Chengdu, P. R. China. Die Arbeit wurde bisher 23-mal zitiert (Crossref) und nennt 38 Literaturangaben. Gefördert von: Natural Science Foundation of China, National Research Foundation of Singapore under its AI Singapore Programme.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Neural Architecture Search (NAS) has recently shown a powerful ability to engineer networks automatically on various tasks. Most current approaches navigate the search direction with the validation performance-based architecture evaluation methodology, which estimates an architecture’s quality by training and validating on a specific large dataset. However, for small-scale datasets, the model’s performance on the validation set cannot precisely estimate that on the test set. The imprecise architecture evaluation can mislead the search to sub-optima. To address the above problem, we propose an efficient multi-objective evolutionary zero-shot NAS framework by evaluating architectures with zero-cost metrics, which can be calculated with randomly initialized models in a training-free manner. Specifically, a general zero-cost metric design principle is proposed to unify the current metrics and help develop several new metrics. Then, we offer an efficient computational method for multi-zero-cost metrics by calculating them in one forward and backward pass. Finally, comprehensive experiments have been conducted on NAS-Bench-201 and MedMNIST. The results have shown that the proposed method can achieve sufficiently accurate, high-throughput performance on MedMNIST and 20[Formula: see text]faster than the previous best method.

Schlagwörter

Stichwörter der Autor:innen

Neural architecture searchimage classificationzero-cost metriczero-shot NAS

Autor:innen und Institutionen

  • College of Computer Science, Sichuan University, Section 4, Southern 1st Ring Road, Chengdu, P. R. China.
  • Institute of High Performance Computing, Agency for Science, Technology and Research (A*STAR), Singapore 138632, Singapore.

Förderung

  • Natural Science Foundation of China
  • National Research Foundation of Singapore under its AI Singapore Programme

Zitieren

APA

Zhang, J., Zhang, L., Wang, Y., Wang, J., Wei, X., & Liu, W. (2023). An Efficient Multi-Objective Evolutionary Zero-Shot Neural Architecture Search Framework for Image Classification. International Journal of Neural Systems, 33(05). https://doi.org/10.1142/s0129065723500168

Vancouver

Zhang J, Zhang L, Wang Y, Wang J, Wei X, Liu W. An Efficient Multi-Objective Evolutionary Zero-Shot Neural Architecture Search Framework for Image Classification. International Journal of Neural Systems. 2023;33(05). doi:10.1142/s0129065723500168

BibTeX
@article{Zhang2023,
  author = {Zhang, Jianwei and Zhang, Lei and Wang, Yan and Wang, Junyou and Wei, Xin and Liu, Wenjie},
  title = {An Efficient Multi-Objective Evolutionary Zero-Shot Neural Architecture Search Framework for Image Classification},
  journal = {International Journal of Neural Systems},
  year = {2023},
  volume = {33},
  number = {05},
  doi = {10.1142/s0129065723500168}
}
Kostenlos starten →