Supervised Stochastic Gradient Algorithms for Multi-Trial Source Separation
Machine Learning
2025-08-29 v1 Machine Learning
Abstract
We develop a stochastic algorithm for independent component analysis that incorporates multi-trial supervision, which is available in many scientific contexts. The method blends a proximal gradient-type algorithm in the space of invertible matrices with joint learning of a prediction model through backpropagation. We illustrate the proposed algorithm on synthetic and real data experiments. In particular, owing to the additional supervision, we observe an increased success rate of the non-convex optimization and the improved interpretability of the independent components.
Cite
@article{arxiv.2508.20618,
title = {Supervised Stochastic Gradient Algorithms for Multi-Trial Source Separation},
author = {Ronak Mehta and Mateus Piovezan Otto and Noah Stanis and Azadeh Yazdan-Shahmorad and Zaid Harchaoui},
journal= {arXiv preprint arXiv:2508.20618},
year = {2025}
}