Spectral clustering: Difference between revisions

Content deleted Content added
Software: started a new section on the costs
Line 57:
| s2cid = 88518155
}}</ref>
 
==Costs==
Denoting the number of the data points ny <math>n</math>, it is important to estimate the memory footprint and compute time, or number of arithmetic operations performed, as a function of <math>n</math>. No matter the algorithm of the spectral clustering, the two main costly items are the construction of the graph Laplacian and determining its eigenvectors for the spectral embedding.
 
==Software==