Enhancing the extrapolative capabilities of deep learning models on temporal parameter sets in femtosecond laser processing
Abstract
Deep learning has emerged as a powerful tool for modeling multi-scale phenomena in laser processing. However, a major challenge remains in achieving reliable extrapolation beyond the range of training data. In this presentation, we propose a neural network architecture designed to improve extrapolative performance for time-domain parameter sets in femtosecond laser processing.
Fields of Study & Subjects
Laser Material Processing Techniques
Laser-Matter Interactions and Applications
Advanced Machining and Optimization Techniques
Publication details
- Journal / source
- Laser-based Micro- and Nanoprocessing
- Pages
- 26–26
- Publication date
- 4 March 2026
- Language
- EN