Noncommutative C^*-algebra Net: Learning Neural Networks with Powerful Product Structure in C^*-algebra

2023/01/26 by Ryuichiro Hataya, Hataya, Ryuichiro, Yuka Hashimoto +1
Computer Science · #Advanced Graph Neural Networks #FOS: Computer and information sciences #FOS: Mathematics #Functional Analysis (math.FA) #Machine Learning (cs.LG) #Machine Learning and Algorithms #Operator Algebras (math.OA) #Stochastic Gradient Optimization Techniques

paper · pdf · doi:10.48550/arxiv.2302.01191

openalex publication_date 2023/01/26 · openalex created_date 2023/02/04 · openalex updated_date 2026/07/28

Abstract

We propose a new generalization of neural network parameter spaces with noncommutative C^*-algebra, which possesses a rich noncommutative structure of products. We show that this noncommutative structure induces powerful effects in learning neural networks. Our framework has a wide range of applications, such as learning multiple related neural networks simultaneously with interactions and learning equivariant features with respect to group actions. Numerical experiments illustrate the validity of our framework and its potential power.

Related

• Convolutional Filtering and Neural Networks with Non Commutative\n Algebras
• Non Commutative Convolutional Signal Models in Neural Networks: Stability to Small Deformations
• Noncommutative Geometry of Computational Models and Uniformization for Framed Quiver Varieties
• Discrete neural nets and polymorphic learning
• Algebra Unveils Deep Learning -- An Invitation to Neuroalgebraic Geometry