Abstract

Spectral clustering transforms the data clustering problem into a graph-partitioning problem and classifies data points by finding the optimal sub-graphs. Traditional spectral clustering algorithms use Gaussian kernel function to construct the similarity matrix, so they are sensitive to the selection of scale parameter. In addition, they need to randomly determine the initial cluster centers at the clustering stage and the clustering performance is not stable. Therefore, this paper presents an algorithm on the basis of message passing, which makes use of a density adaptive similarity measure, describing the relations between data points and obtaining high-quality cluster centers through message passing mechanism in AP clustering. The performance of clustering is optimized by this method. The experiments show that the proposed algorithm can effectively deal with the clustering problem of multi-scale datasets. Moreover, its clustering performance is very stable, and the clustering quality is better than traditional spectral clustering algorithm and k-means algorithm.

Highlights

  • The purpose of clustering is divided data points into different clusters according to their similarity, which makes the similarity of data in the same cluster larger and the similarity of data between different clusters smaller

  • In order to solve the problem of traditional spectral clustering algorithm, which is sensitive to scale parameters, and clustering center initialization, we propose a Spectral Clustering Algorithm Based on Message Passing (MPSC) and use density sensitive similarity to measure the similarity of data points

  • We introduce the ‘‘message passing’’ mechanism in Affinity Propagation (AP) clustering into spectral clustering, which is used to determine the clustering center to improve the performance of spectral clustering algorithm

Read more

Summary

INTRODUCTION

The purpose of clustering is divided data points into different clusters according to their similarity, which makes the similarity of data in the same cluster larger and the similarity of data between different clusters smaller. Spectral clustering algorithm views each point in the data set as the vertex of the graph, and the similarity between any two points as the weight of edges connecting the two vertices. Liu et al [11] obtains the implicit cluster structure features in data by local density, and combines with self-adjusting Gauss kernel function, proposes a spectral clustering algorithm based on shared neighborhoods adaptive similarity. Tao et al [13] puts forward a similarity calculation method between points of manifold structure data to improve the clustering performance of the algorithm. In order to solve the problem of traditional spectral clustering algorithm, which is sensitive to scale parameters, and clustering center initialization, we propose a Spectral Clustering Algorithm Based on Message Passing (MPSC) and use density sensitive similarity to measure the similarity of data points.

INITIALIZATION SENSITIVITY ANALYSIS OF SPECTRAL CLUSTERING ALGORITHM
D1 and and
THE PROPOSED SPECTRAL CLUSTERING BASED ON MESSAGE PASSING APPROACH
Findings
CONCLUSIONS
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call