Programmable Photonic Extreme Learning Machines

Abstract

Photonic neural networks offer a promising alternative to traditional electronic systems for machine learning accelerators due to their low latency and energy efficiency. However, the challenge of implementing the backpropagation algorithm during training has limited their development. To address this, alternative machine learning schemes, such as extreme learning machines (ELMs), are proposed. ELMs use a random hidden layer to increase the feature space dimensionality, requiring only the output layer to be trained through linear regression, thus reducing training complexity. Here, a programmable photonic extreme learning machine (PPELM) is experimentally demonstrated using a hexagonal waveguide mesh, and which enables to program directly on chip the input feature vector and the random hidden layer. This system also permits to apply the nonlinearity directly on-chip by using the system's integrated photodetecting elements. Using the PPELM, three different complex classification tasks are solved successfully. Additionally, two techniques are also proposed and demonstrated to increase the accuracy of the models and reduce their variability using an evolutionary algorithm and a wavelength division multiplexing approach, obtaining excellent performance. These results show that programmable photonic processors may become a feasible way to train competitive machine learning models on a versatile and compact platform.