Now 8413 visitors
Today:0 Yesterday:0
Total: 162435 216S 51P 5R
2026-10-22, Week 43
Member Login
TACT Journal Homepage
Call for Paper
Committee
Find My Paper
Author Homepage
Statistics & History
Welcome Message
Paper Procedure
FAQ
Paper Archives
Outstanding Papers
Program & Proceedings
Presentation Platform
Hotel & Travel Info
Photo Gallery
Scheduler Login
Seminar
Archives Login
Sponsors


















Virual Oral Question and Answer
IEEE/ICACT20240208 Question.2
Questioner: jllovido@bicol-u.edu.ph    2024-02-08 오전 12:19:36 		IEEE/ICACT20240208 Answer.2
Answer by Auhor jjhu@nttu.edu.tw   2024-02-08 오전 12:19:36
Are there any potential limitations or areas for further improvement identified with the proposed PINN method? We believe that the potential limitations and areas for further improvement with physics-informed neural networks are: Generalization: PINNs can sometimes overfit to specific training data and fail to generalize to new scenarios. High computational cost: Training PINNs typically requires many significant computing resources. Further work on optimization methods could help address this. Robustness: It's not clear yet how sensitive PINNs are to noise or uncertainty in the data and governing equations. Complex physics: Current PINNs have focused mainly on relatively simple physical systems described by partial differential equations. Extending to more complex physics with coupled multi-physics systems remains an open challenge.

Select Voice