Duality and DeepMartingale for High-Dimensional Optimal Switching: Computable Upper Bounds and Approximation-Expressivity Guarantees
Junyan Ye, Hoi Ying Wong · 2026
Read Duality and DeepMartingale for High-Dimensional Optimal Switching: Computable Upper Bounds and Approximation-Expressivity Guarantees by Junyan Ye, Hoi Ying Wong (2026) 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-2604-08080.txt.
Description
Ye and Wong study optimal switching over finite horizons, with set intervention dates and continuous-time observations between decisions. They build a dual deep-learning framework yielding computable upper bounds. A dual representation for multiple switching comes from martingale penalties, with the smallest given by Doob martingales of continuation values, so the upper bound is fully known. They carry DeepMartingale from optimal stopping to optimal switching, proving convergence for the upper-bound loss and an L2-surrogate loss. An expressivity analysis follows under the stated structural assumptions.
Catalog record
| CREATOR | Junyan Ye, Hoi Ying Wong |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 100613 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.08080