Overview
The Journal of Inverse and Ill-posed Problems is inviting submissions for a special issue focused on “Physics-Informed Machine Learning for Problems in Science.” This initiative aims to explore the rapidly evolving field of utilizing physics-informed machine learning (PIML) to tackle inverse problems across various scientific domains. The significance of this special issue lies in its potential to advance methodologies that integrate domain knowledge into machine learning frameworks, thereby enhancing the accuracy and reliability of solutions.
Background & Relevance
Inverse problems are prevalent in numerous scientific fields, including material science, geophysics, and biomedical engineering. These problems involve deducing unknown causes from observable effects, often presenting challenges due to their ill-posed nature. The incorporation of physics-informed approaches can provide essential regularization, improving the robustness of solutions. Understanding and addressing these challenges is crucial for advancing scientific inquiry and practical applications in engineering and technology.
Key Details
- Submission Deadline: October 1, 2025
- Expected Publication Date: December 31, 2025
- Topics of Interest:
- System parameters/equations identification
- Data assimilation
- History matching on measurements
- Building digital twins of complex systems
- Uncertainty estimation
- Construction of physics-informed forward surrogate models
- Guest Editors:
- Prof. Evgeny Burnaev (Skoltech, Russia)
- Prof. N.M. Anoop Krishnan (IIT Delhi, India)
- Prof. Ivan Oseledets (AIRI, Russia)
Eligibility & Participation
This call for papers targets researchers and practitioners in the fields of machine learning, physics, and engineering who are working on or interested in the application of physics-informed techniques to solve inverse problems. Contributions that advance theoretical foundations, computational frameworks, and real-world implementations are particularly encouraged.
Submission or Application Guidelines
- Review the submission information at Journal Submission Guidelines.
- Submit your manuscript through the journal’s webpage at Manuscript Submission. Clearly indicate that your submission is for the special issue on the first page of your manuscript and in the cover letter.
- Notify the leading editor, Prof. Evgeny Burnaev, via email at E.Burnaev@skoltech.ru with the subject line “JIIP special issue PIML submission.”
- Early submissions are encouraged, and the review process will commence upon receipt of contributions.
More Information
This special issue aims to highlight innovative methodologies that address the fundamental challenges in physics-informed machine learning for inverse problems. By showcasing transformative applications across various scientific domains, it seeks to foster collaboration and knowledge sharing among researchers. The emphasis on reproducibility and the availability of benchmarks will enhance the credibility and impact of the published works.
Conclusion
Researchers are encouraged to contribute their insights and findings to this special issue, which promises to be a significant platform for advancing the field of physics-informed machine learning. Explore this opportunity to share your work and engage with the broader scientific community.
Category: CFP & Deadlines
Tags: physics-informed machine learning, inverse problems, data assimilation, uncertainty estimation, geophysics, fluid dynamics, biomedical imaging, digital twins, machine learning, open-source, neural networks, computational frameworks
