On a Hierarchy of Spectral Invariants for Graphs

V. Arvind,Frank Fuhlbrück,Johannes Köbler,Oleg Verbitsky

2024-05-30

Abstract:We consider a hierarchy of graph invariants that naturally extends the spectral invariants defined by Fürer (Lin. Alg. Appl. 2010) based on the angles formed by the set of standard basis vectors and their projections onto eigenspaces of the adjacency matrix. We provide a purely combinatorial characterization of this hierarchy in terms of the walk counts. This allows us to give a complete answer to Fürer's question about the strength of his invariants in distinguishing non-isomorphic graphs in comparison to the 2-dimensional Weisfeiler-Leman algorithm, extending the recent work of Rattan and Seppelt (SODA 2023). As another application of the characterization, we prove that almost all graphs are determined up to isomorphism in terms of the spectrum and the angles, which is of interest in view of the long-standing open problem whether almost all graphs are determined by their eigenvalues alone. Finally, we describe the exact relationship between the hierarchy and the Weisfeiler-Leman algorithms for small dimensions, as also some other important spectral characteristics of a graph such as the generalized and the main spectra.

Computational Complexity,Combinatorics

What problem does this paper attempt to address?

The problems that this paper attempts to solve mainly focus on spectral invariants in graph theory and their relationship with combinatorial methods. Specifically, the paper aims to systematically explore the connection between spectral methods and combinatorial methods in finding highly expressive graph invariants. The core problems of the paper can be summarized as follows:

1. **Hierarchy of Spectral Invariants**:
- The paper proposes a hierarchy of graph invariants based on spectral parameters, which naturally extends the spectral invariants defined by Fürer (in 2010). These spectral invariants are based on the angles formed by standard basis vectors and their projections on the eigenspaces of the adjacency matrix.

2. **Answer to Fürer's Problem**:
- The paper provides a purely combinatorial characterization to answer Fürer's question regarding the ability of his spectral invariants to distinguish non - isomorphic graphs. In particular, the paper compares Fürer's spectral invariants with the 2 - dimensional Weisfeiler - Leman algorithm (2 - WL) in terms of their ability to distinguish non - isomorphic graphs.

3. **Are Almost All Graphs Determined by Spectrum and Angles?**:
- The paper proves that almost all graphs can be determined up to isomorphism by spectrum and angles. This problem has been a long - standing unsolved problem in graph theory, that is, whether almost all graphs can be determined solely by their eigenvalues.

4. **Relationship between Spectral Invariants and the Weisfeiler - Leman Algorithm**:
- The paper describes in detail the exact relationship between the hierarchy of spectral invariants and the Weisfeiler - Leman algorithm in small dimensions, and explores some important spectral properties, such as the generalized spectrum and the principal spectrum.

### Main Contributions

- **Pure Combinatorial Characterization**: The paper provides a pure combinatorial method to characterize the hierarchy of spectral invariants, which makes it possible to understand these invariants directly from path counting in graphs without relying on linear algebra.
- **Complete Answer to Fürer's Problem**: The paper not only answers the question raised by Fürer, but also further proves that almost all graphs can be determined up to isomorphism by weak spectral invariants (weak - FSI).
- **New Separation Results**: The paper shows the strict relationships in the hierarchy of spectral invariants and provides specific pairs of graphs to separate different invariants, such as strong - FSI and WL3/2.

### Conclusion

By systematically studying the relationship between spectral invariants and combinatorial methods, the paper not only solves the problems raised by Fürer, but also provides new perspectives and tools for the study of the graph isomorphism problem. These results are of great significance for understanding the structure and properties of graphs, especially in the fields of computer science and graph theory.