EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Uniform Loss Versus Specialized Optimization: A Comparative Analysis in Multi-Task Learning

Gabriel S. Gama, Valdir Grassi

IEEE Access · 2026 · 14 · 44079-44094

DOI: 10.1109/access.2026.3676354 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Gabriel S. Gama und 1 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior, Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Specialized Multi-Task Optimizers (SMTOs) balance task learning in Multi-Task Learning by addressing issues like conflicting gradients and differing gradient norms, which hinder equal-weighted task training. However, recent critiques suggest that equally weighted tasks can achieve competitive results compared to SMTOs, arguing that previous SMTO results were influenced by poor hyperparameter optimization and lack of regularization. In this work, we evaluate these claims through an extensive empirical evaluation of SMTOs, including some of the latest methods, on more complex multi-task problems to clarify this behavior. More specifically, we start our analysis by evaluating all SMTOs on a simple MNIST problem to identify the promising optimizers and assess them further on progressively more complex multi-task problems. Our findings indicate that SMTOs perform well compared to uniform loss and that fixed weights can achieve competitive performance compared to SMTOs. Furthermore, we demonstrate why uniform loss performs similarly to SMTOs in some instances. The source code is available at https://github.com/Gabriel-SGama/UnitScal_vs_SMTOs

Förderung

  • Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior
  • Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico

Zitieren

APA

Gama, G. S., & Grassi, V. (2026). Uniform Loss Versus Specialized Optimization: A Comparative Analysis in Multi-Task Learning. IEEE Access, 14, 44079-44094. https://doi.org/10.1109/access.2026.3676354

Vancouver

Gama GS, Grassi V. Uniform Loss Versus Specialized Optimization: A Comparative Analysis in Multi-Task Learning. IEEE Access. 2026;14:44079-44094. doi:10.1109/access.2026.3676354

BibTeX
@article{Gama2026,
  author = {Gama, Gabriel S. and Grassi, Valdir},
  title = {Uniform Loss Versus Specialized Optimization: A Comparative Analysis in Multi-Task Learning},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {44079--44094},
  doi = {10.1109/access.2026.3676354}
}
Kostenlos starten →