说明：  自适应谱聚类研究的论文，可供科研人员参考
(Spectral clustering hasbeenusedto identify clustersthat arenon
linearly separableininput space, andusuallyoutper
formstraditional clustering algorithms. However, the performances of spectral clustering are severely dependent onvaluesof the scaling parameter. Inthispaper, anadaptive spectral clustering(ASC) algorithmwasproposedbasedontraditional spectral clus
tering, whichcanchoose the scaling parameter automatically by using techniques similar to kernel selection. The newalgorithm was comparedto existingparameter selectionbasedspectral clustering algorithmsonbothsyntheticandUCI datasets, andthe ex
perimental results validatethe effectivenessof theproposedalgorithm)