Comparing Machine Learning Methods on Synthetic Laser Accelerated Proton Data

The rapid advancements in laser systems have enabled high-speed data acquisition with kHz repetition rates. However, due to the complex nature of laser–matter interactions and the limitations of analytical and computational methods, accurately characterizing their features remains challenging and costly. In this research, we harness the potential of machine learning to emulate laser interactions using synthetic datasets. By comparing the performance of neural networks, decision trees, and random forests, we identify effective and efficient approaches for analyzing real-world laser datasets generated by emerging laser systems.