# Spectral Clustering Methods for Multiplex Networks

## TL;DR

This paper extends spectral clustering techniques to multiplex networks, addressing challenges in defining natural generalizations and comparing models on synthetic data to identify effective community detection methods.

## Contribution

It introduces a spectral clustering extension for multiplex networks and evaluates different models, highlighting the superiority of a dynamically motivated approach.

## Key findings

- Dynamically motivated models outperform structurally motivated ones in community detection.
- Spectral clustering can be adapted to multiplex networks with specific considerations.
- Synthetic network experiments demonstrate the effectiveness of the proposed methods.

## Abstract

Multiplex networks offer an important tool for the study of complex systems and extending techniques originally designed for single--layer networks is an important area of study. One of the most important methods for analyzing networks is clustering the nodes into communities that represent common connectivity patterns. In this paper we extend spectral clustering to multiplex structures and discuss some of the difficulties that arise in attempting to define a natural generalization. In order to analyze our approach, we describe three simple, synthetic multiplex networks and compare the performance of different multiplex models. Our results suggest that a dynamically motivated model is more successful than a structurally motivated model in discovering the appropriate communities.