Predicting Supercontinuum Generation Dynamics Using a Neural Network

The ultrabroadband supercontinuum (SC) source possesses a number of unique characteristics which have had a major impact in many research fields, ranging from precision metrology to high resolution imaging and ultrafast spectroscopy [1] . The generation of a supercontinuum is a highly complex process resulting from the propagation of intense light pulses in a nonlinear medium such as an optical fibre or waveguide. As a result of the very rich landscape of nonlinear dynamics, the input pulses injected into the fibre experience massive spectral broadening that can expand an initially narrowband input field to an output spanning from the visible to infrared [1] .