A clustering method based on quantum mechanics was proposed to analyze data clustering in pattern recognition problems. Quantum clustering (QC) was used to construct potential function for global spherical symmetry on the basis of data points controlling width of structures. The proposed algorithm consisted of single parameter to determine the scale over which the structures were searched. Analysis suggested that the method could be efficiently applied in higher dimensions by limiting the evaluation of Schrodinger potential to location of data points.
|Number of pages||4|
|Journal||Physical Review Letters|
|State||Published - 7 Jan 2002|