Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.02439 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910038896934912 |
|---|---|
| author | Hammond, Joshua E. Soderstrom, Tyler A. Korgel, Brian A. Baldea, Michael |
| author_facet | Hammond, Joshua E. Soderstrom, Tyler A. Korgel, Brian A. Baldea, Michael |
| contents | Data-driven models of dynamical systems require extensive amounts of training data. For many practical applications, gathering sufficient data is not feasible due to cost or safety concerns. This work uses the Subset Extended Kalman Filter (SEKF) to adapt pre-trained neural network models to new, similar systems with limited data available. Experimental validation across damped spring and continuous stirred-tank reactor systems demonstrates that small parameter perturbations to the initial model capture target system dynamics while requiring as little as 1% of original training data. In addition, finetuning requires less computational cost and reduces generalization error. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_02439 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Using the SEKF to Transfer NN Models of Dynamical Systems with Limited Data Hammond, Joshua E. Soderstrom, Tyler A. Korgel, Brian A. Baldea, Michael Machine Learning Data-driven models of dynamical systems require extensive amounts of training data. For many practical applications, gathering sufficient data is not feasible due to cost or safety concerns. This work uses the Subset Extended Kalman Filter (SEKF) to adapt pre-trained neural network models to new, similar systems with limited data available. Experimental validation across damped spring and continuous stirred-tank reactor systems demonstrates that small parameter perturbations to the initial model capture target system dynamics while requiring as little as 1% of original training data. In addition, finetuning requires less computational cost and reduces generalization error. |
| title | Using the SEKF to Transfer NN Models of Dynamical Systems with Limited Data |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2603.02439 |