HomeResearchSparse Subspace Clustering for Big Data
1 Motivation With the rapid increase of data amount, a dataset is always composed of huge amount of data points. To reduce the computational cost in such big data problems, it is crucial to pick out and only deal with…
1 Motivation Distinguished from the traditional clustering methods based on Euclidean distance, subspace clustering methods assign the same label to data points lying in the same subspace. As the Euclidean distance between length limited data points would become approximately the…