Dirac-Equation Signal Processing: Physics Boosts Topological Machine Learning
arXiv:2412.05132 · doi:10.1093/pnasnexus/pgaf139
Abstract
Topological signals are variables or features associated with both nodes and edges of a network. Recently, in the context of Topological Machine Learning, great attention has been devoted to signal processing of such topological signals. Most of the previous topological signal processing algorithms treat node and edge signals separately and work under the hypothesis that the true signal is smooth and/or well approximated by a harmonic eigenvector of the Hodge-Laplacian, which may be violated in practice. Here we propose Dirac-equation signal processing, a framework for efficiently reconstructing true signals on nodes and edges, also if they are not smooth or harmonic, by processing them jointly. The proposed physics-inspired algorithm is based on the spectral properties of the topological Dirac operator. It leverages the mathematical structure of the topological Dirac equation to boost the performance of the signal processing algorithm. We discuss how the relativistic dispersion relation obeyed by the topological Dirac equation can be used to assess the quality of the signal reconstruction. Finally, we demonstrate the improved performance of the algorithm with respect to previous algorithms. Specifically, we show that Dirac-equation signal processing can also be used efficiently if the true signal is a non-trivial linear combination of more than one eigenstate of the Dirac equation, as it generally occurs for real signals.
(14 pages, 7 figures)
References in corpus (35)
- Geometric deep learning: going beyond Euclidean data
- Machine learning and the physical sciences
- Networks beyond pairwise interactions: structure and dynamics
- The physics of higher-order interactions in complex systems
- A roadmap for the computation of persistent homology
- Clique topology reveals intrinsic geometric structure in neural correlations
- What are higher-order networks?
- TopologyNet: Topology based deep convolutional neural networks for biomolecular property predictions
- Explosive higher-order Kuramoto dynamics on simplicial complexes
- Topological Signal Processing over Simplicial Complexes
- Topological strata of weighted complex networks
- Random Walks on Simplicial Complexes and the normalized Hodge 1-Laplacian
- Emergent Hyperbolic Network Geometry
- Simplicial complexes and complex systems
- Signal Processing on Higher-Order Networks: Livin' on the Edge ... and Beyond
- Network geometry with flavor: from complexity to quantum geometry
- Unveiling the higher-order organization of multivariate time series
- Simplicial complexes: higher-order spectral dimension and dynamics
- Weighted simplicial complexes and their representation power of higher-order network data and topology
- The topological Dirac equation of networks and simplicial complexes
- Persistent Dirac for molecular representation
- Cell Complex Neural Networks
- Dirac signal processing of higher-order topological signals
- Balanced Hodge Laplacians Optimize Consensus Dynamics over Simplicial Complexes
- Dirac synchronization is rhythmic and explosive
- Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs
- Mesoscale analyses of fungal networks as an approach for quantifying phenotypic traits
- Diffusion-driven instability of topological signals coupled by the Dirac operator
- Topological Deep Learning: Going Beyond Graph Data
- Local Dirac Synchronization on Networks
- The mass of simple and higher-order networks
- Architectures of Topological Deep Learning: A Survey of Message-Passing Topological Neural Networks
- Global topological synchronization of weighted simplicial complexes
- Quantum entropy couples matter with geometry
- Sheaf Hypergraph Networks