Convolutional neural network-based modeling of an ultrafast laser

We continue to see massive developments in the synergistic fields of particle accelerators and lasers. These beam sources have proven over and over to be essential tools for scientists, engineers, and technologists in many fields, including discovery science, medicine, environmental applications, energy application, and industry. Machine learning (ML) techniques to help the model and control of the laser beam for driving particle accelerators. We know from our experiences and the experiences of others that using advanced control techniques, including ML can often improve processes and machine performance. Our research is based on Convolutional-Neural-Network; as one of the classes of neural-networks. In this research, we report the Neural Network (NN) based on the feedforward-backpropagation control system to model the relationship between the DAZZLER inputs and the temporal pulse width of the femtosecond pulsed laser system.