Menu
Home
People
Places
Arts
History
Plants & Animals
Science
Life & Culture
Technology
Reference.org
Spectral clustering
open-in-new
<p>In <a href="/facts/Multivariate_statistics/tDuNcWrL">multivariate statistics</a>, spectral clustering techniques make use of the <a href="/facts/Spectrum_of_a_matrix/BIBnA4qf">spectrum</a> (<a href="/facts/Eigenvalues/8TjEoT8u">eigenvalues</a>) of the <a href="/facts/Similarity_matrix/FmmFco8c">similarity matrix</a> of the data to perform <a href="/facts/Dimensionality_reduction/XoRpcrB0">dimensionality reduction</a> before <a href="/facts/Cluster_analysis/HUeDnWPt">clustering</a> in fewer dimensions. The similarity matrix is provided as an input and consists of a quantitative assessment of the relative similarity of each pair of points in the dataset. </p><p>In application to image segmentation, spectral clustering is known as <a href="/facts/Segmentation-based_object_categorization/Wp6XnTx0">segmentation-based object categorization</a>. </p>