Machine Learning and Invariant Theory

In this project we explore the use of invariant theory on machine learning.  

Topics

Some of the topics that we are interested are:

• Group actions and machine learning

• Convolutional Neural Networks

Conferences

• Artificial Intelligence in Mathematics, AMS Meeting #1195, University of Wisconsin-Milwaukee, Milwaukee, WI, April 20-21, 2024. 

Links

• VC dimensions of group convolutional neural networks

• Topology-aware Convolutional Neural Network for Efficient Skeleton-based Action Recognition

• MVC-Net: A Convolutional Neural Network Architecture for Manifold-Valued Images With Applications

• Clebsch-Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network

• On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups

• Introduction to Equivariant Machine Learning S.Villar