Addressing Data Scarcity in Optical Matrix Multiplier Modeling Using Transfer Learning

TL;DR

This paper introduces a transfer learning approach that combines synthetic and experimental data to improve neural network modeling of optical matrix multipliers, significantly reducing errors with limited data.

Contribution

The study demonstrates a novel transfer learning method that enhances optical neural network modeling accuracy using minimal experimental data.

Findings

Achieved < 1 dB RMS error with only 25% of data

Transfer learning outperforms standalone models and analytical models

Regularization and ensemble techniques improve accuracy

Abstract

We present and experimentally evaluate using transfer learning to address experimental data scarcity when training neural network (NN) models for Mach-Zehnder interferometer mesh-based optical matrix multipliers. Our approach involves pre-training the model using synthetic data generated from a less accurate analytical model and fine-tuning with experimental data. Our investigation demonstrates that this method yields significant reductions in modeling errors compared to using an analytical model, or a standalone NN model when training data is limited. Utilizing regularization techniques and ensemble averaging, we achieve < 1 dB root-mean-square error on the matrix weights implemented by a 3x3 photonic chip while using only 25% of the available data.