Locating Eigenvalues of a Symmetric Matrix whose Graph is Unicyclic

We present a linear-time algorithm that computes in a given real interval the number of eigenvalues of any symmetric matrix whose underlying graph is unicyclic. The algorithm can be applied to vertex- and/or edge-weighted or unweighted unicyclic graphs. We apply the algorithm to obtain some general results on the spectrum of a generalized sun graph for certain matrix representations which include the Laplacian, normalized Laplacian and signless Laplacian matrices.

Multilabel Classification Based on Graph Neural Networks

Typical Laplacian embedding focuses on building Laplacian matrices prior to minimizing weights of connected graph components. However, for multilabel problems, it is difficult to determine such Laplacian graphs owing to multiple relations between vertices. Unlike typical approaches that require precomputed Laplacian matrices, this chapter presents a new method for automatically constructing Laplacian graphs during Laplacian embedding. By using trace minimization techniques, the topology of the Laplacian graph can be learned from input data, subsequently creating robust Laplacian embedding and influencing graph convolutional networks. Experiments on different open datasets with clean data and Gaussian noise were carried out. The noise level ranged from 6% to 12% of the maximum value of each dataset. Eleven different multilabel classification algorithms were used as the baselines for comparison. To verify the performance, three evaluation metrics specific to multilabel learning are proposed because multilabel learning is much more complicated than traditional single-label settings; each sample can be associated with multiple labels. The experimental results show that the proposed method performed better than the baselines, even when the data were contaminated by noise. The findings indicate that the proposed method is reliably robust against noise.

THE DISTANCE RELATED SPECTRA OF SOME SUBDIVISION RELATED GRAPHS

Let $G$ be a connected graph with a distance matrix $D$. The distance eigenvalues of $G$ are the eigenvalues of $D$, and the distance energy $E_D(G)$ is the sum of its absolute values. The transmission $Tr(v)$ of a vertex $v$ is the sum of the distances from $v$ to all other vertices in $G$. The transmission matrix $Tr(G)$ of $G$ is a diagonal matrix with diagonal entries equal to the transmissions of vertices. The matrices $D^L(G)= Tr(G)-D(G)$ and $D^Q(G)=Tr(G)+D(G)$ are, respectively, the Distance Laplacian and the Distance Signless Laplacian matrices of $G$. The eigenvalues of $D^L(G)$ ( $D^Q(G)$) constitute the Distance Laplacian spectrum ( Distance Signless Laplacian spectrum ). The subdivision graph $S(G)$ of $G$ is obtained by inserting a new vertex into every edge of $G$. We describe here the Distance Spectrum, Distance Laplacian spectrum and Distance Signless Laplacian spectrum of some types of subdivision related graphs of a regular graph in the terms of its adjacency spectrum. We also derive analytic expressions for the distance energy of $\bar{S}(C_p)$, partial complement of the subdivision of a cycle $C_p$ and that of $\overline {S\left( {C_p }\right)}$, complement of the even cycle $C_{2p}$.

Computing the Permanent of the Laplacian Matrices of Nonbipartite Graphs

Let G be a graph with Laplacian matrix L G . Denote by per L G the permanent of L G . In this study, we investigate the problem of computing the permanent of the Laplacian matrix of nonbipartite graphs. We show that the permanent of the Laplacian matrix of some classes of nonbipartite graphs can be formulated as the composite of the determinants of two matrices related to those Laplacian matrices. In addition, some recursion formulas on per L G are deduced.

Spectral Clustering of Psychological Networks

Spectral clustering is a well-known method for clustering the vertices of an undirected network. Although its use in network psychometrics has been limited, spectral clustering has a close relationship to the commonly-used walktrap algorithm. In this paper, we report results from four simulation experiments designed to evaluate the ability of spectral clustering and the walktrap algorithm to recover underlying cluster structure in networks. The salient findings include: (1) the cluster recovery performance of the walktrap algorithm can be improved slightly by using exact K-means clustering instead of hierarchical clustering; (2) K-means and K-median clustering led to comparable recovery performance when used to cluster vertices based on the eigenvectors of Laplacian matrices in spectral clustering; (3) spectral clustering using the unnormalized Laplacian matrix generally yielded inferior cluster recovery in comparison to the other methods; (4) when the correct number of clusters was provided for the methods, spectral clustering using the normalized Laplacian matrix led to better recovery than the walktrap algorithm; (5) when the number of clusters was unknown, spectral clustering using the normalized Laplacian matrix was appreciably better than the walktrap algorithm when the clusters were equally-sized, but the two methods were competitive when the clusters were not equally-sized. Overall, both the walktrap algorithm and spectral clustering of the normalized Laplacian matrix are effective for partitioning the vertices of undirected networks, with the latter performing better in most instances.

Linear Software Models: An Occam’s Razor Set of Algebraic Connectors Integrates Modules into a Whole Software System

Well-designed software systems, with providers only modules, have been rigorously obtained by algebraic procedures from the software Laplacian Matrices or their respective Modularity Matrices. However, a complete view of the whole software system should display, besides provider relationships, also consumer relationships. Consumers may have two different roles in a system: either internal or external to modules. Composite modules, including both providers and internal consumers, are obtained from the joint providers and consumers Laplacian matrix, by the same spectral method which obtained providers only modules. The composite modules are integrated into a whole Software System by algebraic connectors. These algebraic connectors are a minimal Occam’s razor set of consumers external to composite modules, revealed through iterative splitting of the Laplacian matrix by Fiedler eigenvectors. The composite modules, of the respective standard Modularity Matrix for the whole software system, also obey linear independence of their constituent vectors, and display block-diagonality. The spectral method leading to composite modules and their algebraic connectors is illustrated by case studies. The essential novelty of this work resides in the minimal Occam’s razor set of algebraic connectors — another facet of Brooks’ Propriety principle leading to Conceptual Integrity of the whole Software System — within Linear Software Models, the unified algebraic theory of software modularity.