Title :
Introduction to spectral clustering
Author :
Hamad, Denis ; Biela, Philippe
Author_Institution :
LASL/ULCO, Calais
Abstract :
Spectral clustering methods are based on graph and matrix theories. Their principle is simple: given some data inputs, build similarity matrix, analyse the spectrum of its Laplacian matrix, and often get a perfect clustering from the eigenvectors analysis. This paper presents an introduction to spectral clustering methods and some applications in signal and image segmentation.
Keywords :
eigenvalues and eigenfunctions; graph theory; matrix algebra; pattern clustering; Laplacian matrix; eigenvectors analysis; graph theory; matrix theory; similarity matrix; spectral clustering; Application software; Clustering algorithms; Clustering methods; Computer vision; Eigenvalues and eigenfunctions; Image segmentation; Laplace equations; Machine learning; Partitioning algorithms; Signal processing algorithms; Spectral methods; clustering; eigenvalue; graph; partition; segmentation;
Conference_Titel :
Information and Communication Technologies: From Theory to Applications, 2008. ICTTA 2008. 3rd International Conference on
Conference_Location :
Damascus
Print_ISBN :
978-1-4244-1751-3
Electronic_ISBN :
978-1-4244-1752-0
DOI :
10.1109/ICTTA.2008.4529994
