Bi-Level optimization for parameter estimation of differential equations using interpolation
Siddharth Prabhu, Srinivas Rangarajan, Mayuresh Kothare · 2025
Read Bi-Level optimization for parameter estimation of differential equations using interpolation by Siddharth Prabhu, Srinivas Rangarajan, Mayuresh Kothare (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-2506-00720.txt.
Description
Siddharth Prabhu, Srinivas Rangarajan, and Mayuresh Kothare cut the cost of fitting differential equation models to data. To estimate parameters for ordinary differential equations means to cut the gap between model and experiment again and again. The famed single-shooting method integrates the equations at each step. To compute gradients calls for sensitivity derivatives through the integrator, which is costly.
The team uses interpolation to cut the cost of those sensitivity calculations. On that base, they propose a bi-level optimization frame. It uses the structure of the differential equations and solves a convex inner problem. They use the frame for standard parameter gauges and for data-driven dynamic model discovery. The path gauges correct parameters on benchmark problems. It also extends with ease to delay, stiff, and partly seen equations with no large changes.
Catalog record
| CREATOR | Siddharth Prabhu, Srinivas Rangarajan, Mayuresh Kothare |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-00720