Saturated linear unit as an universal symmetric activation function for deep learning
Serbian Journal of Management · 2026 · 21 (1) · 37-48
DOI: 10.5937/sjm21-57891 ↗Kurz gefasst
Fachartikel in der Zeitschrift Serbian Journal of Management, 2026. Verfasst von Banduka Lutovac und 4 weiteren Autor:innen. Die Arbeit nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Education, Science and Technological Development of the Republic of Serbia.
Abstract (Originalsprache)
There is a number of symmetric activation functions used in artificial neural networks for deep learning. In this paper, we propose a universal activation function based on the Saturated Linear Unit (SaLU) that can be used instead of any known symmetric activation function. It is not necessary for classification tasks to have an exact calculation of the probability of detected classes. The classification decision is made based on the highest probability for the input values. We propose, as a proof of concept, that the two most commonly used hyperbolic tangent and algebraic sigmoid activation functions can be effectively replaced by SaLU by choosing a single parameter. Moreover, the theoretical step function can also be replaced by SaLU for a wider transition range. All derivations use symbolic processing. Also shown is a visualization of the range of inputs that result in a suitable classification. This can help scientists and programmers design complex machine learning algorithms and understand how deep learning algorithms work.
Förderung
- Ministry of Education, Science and Technological Development of the Republic of Serbia
Zitieren
APA
Lutovac, B., Milićević, V., Franc, I., Zdravković, N., & Dimitrijević, N. (2026). Saturated linear unit as an universal symmetric activation function for deep learning. Serbian Journal of Management, 21(1), 37-48. https://doi.org/10.5937/sjm21-57891
Vancouver
Lutovac B, Milićević V, Franc I, Zdravković N, Dimitrijević N. Saturated linear unit as an universal symmetric activation function for deep learning. Serbian Journal of Management. 2026;21(1):37-48. doi:10.5937/sjm21-57891
BibTeX
@article{Lutovac2026,
author = {Lutovac, Banduka and Milićević, Vladimir and Franc, Igor and Zdravković, Nemanja and Dimitrijević, Nikola},
title = {Saturated linear unit as an universal symmetric activation function for deep learning},
journal = {Serbian Journal of Management},
year = {2026},
volume = {21},
number = {1},
pages = {37--48},
doi = {10.5937/sjm21-57891}
}
EUVIMED