An Adaptive Order Caputo Fractional Gradient Descent Method for Multi-objective Optimization Problems
Barsha Shaw, Md Abu Talhamainuddin Ansary · 2025
Read An Adaptive Order Caputo Fractional Gradient Descent Method for Multi-objective Optimization Problems by Barsha Shaw, Md Abu Talhamainuddin Ansary (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-07674.txt.
Description
Barsha Shaw and Md Abu Talhamainuddin Ansary proposed a descent method for problems with several goals at once, in a 2025 paper. Such problems, called multi-objective, need a point where no goal can improve without hurting another. Their method uses a fractional style of slope, named after Caputo, whose order adapts as the search runs.
At each step, the method solves a small subproblem to pick a downhill direction for all goals jointly. A line search in the Armijo style then sets the step length. The method needs no preset parameters and no ranking of the goals in advance. It works on smooth problems and on rough ones alike. The authors prove it settles toward a regularized solution under mild assumptions. They test it on sample problems, including neural network training.
Catalog record
| CREATOR | Barsha Shaw, Md Abu Talhamainuddin Ansary |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-07674