Multi-objective optimization for photonic systems with advanced functionalities and improved performance

The development of photonic system realizing advanced functionalities on-chip requires the careful integration of a large number of reliable and high-performance components. Novel approaches such as non-trivial geometries and metamaterials are required to achieve these targets. As a consequence, new design tools capable of handling a large number of correlated parameters are required. Moreover, multiple figures of merit must be considered simultaneously to evaluate the performance of the selected devices end unsure appropriate system integration. Here, we will discuss the potentiality offered by the combination of machine learning dimensionality reduction and optimization in tackling the multi-objective design of photonic devices as well as for the investigation of the effect of fabrication tolerances.