Tailoring emission spectra by using core-shell nanostructures with a random forest algorithm

Abstract We propose a scheme to tailor emission spectra through the engineering of the local density of states by using core-shell nanostructures. A random forest algorithm is used to train a forward model for spectrum prediction and an inverse model for the construction of core-shell nanostructures. From the simulation results, it can be seen that we can predict the spectra very well and achieve good effects for tailoring the emission spectra with core-shell nanostructures based on the established models, eliminating the time-consuming and laborious design process with manual intervention.