A Sequential Meta-Transfer (SMT) Learning to Combat Complexities of Physics-Informed Neural Networks: Application to Composites Autoclave Processing
Physics-Informed Neural Networks (PINNs) have gained popularity in solving nonlinear partial differential equations (PDEs) via integrating physical laws into the train...
Tags：Paper and LLMsTransfer LearningPricing Type
- Pricing Type: Free
- Price Range Start($):
GitHub Link
The GitHub link is https://github.com/miladramzy/sequentialmetatransferpinns
Introduce
This GitHub repository showcases a JAX implementation of the paper titled “Meta-Transfer Sequential Learning of Physics-Informed Neural Networks in Advanced Composites Manufacturing.” The approach combines sequential learning and meta-transfer learning to enhance the training of Physics-Informed Neural Networks (PINNs) in intricate and nonlinear systems, particularly in advanced composites manufacturing. The framework introduces two sequential learning methods time marching and backward-compatibility. For more details, refer to the paper on arXiv (https://arxiv.org/abs/2308.06447).
Physics-Informed Neural Networks (PINNs) have gained popularity in solving nonlinear partial differential equations (PDEs) via integrating physical laws into the training of neural networks, making them superior in many scientific and engineering applications.
Content
This repository presents a JAX implementation of the paper entitled “MA Sequential Meta-Transfer (SMT) Learning to Combat Complexities of Physics-Informed Neural Networks: Application to Composites Autoclave Processing”. The proposed framework integrates a sequential learning strategy with the meta-transfer learning approach to make the training of PINNs in complex and highly nonlinear system more efficient and adaptable.
Related
