UDuo: Universal Dual Optimization Framework for Online Matching
Bin Li, Diwei Liu, Zehong Hu, Jia Jia · 2025
Read UDuo: Universal Dual Optimization Framework for Online Matching by Bin Li, Diwei Liu, Zehong Hu, Jia Jia (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-2505-22243.txt.
Description
Bin Li, Diwei Liu, Zehong Hu, and Jia Jia propose UDuo, a framework for matching users to resources online under budget limits, in a 2025 paper. Classic methods assume users arrive by a fixed random pattern, an assumption that breaks when real traffic shifts over time. UDuo instead tracks a time-stamped profile of arrivals and spending, learns pacing rules that adapt across constraint setups, and forecasts how arrivals will move next.
Tests on real pricing workloads showed faster convergence and better efficiency than the classic stochastic model. The authors also prove that the method stays near optimal while respecting constraints as conditions change. The framework targets ad systems and other allocation problems where demand will not sit still.
Catalog record
| CREATOR | Bin Li, Diwei Liu, Zehong Hu, Jia Jia |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-22243