lecture12-clustering-handout-6-per

Gk i di ck2 sum over all di in cluster k

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: erms of similari*es) 3 Introduc)on to Informa)on Retrieval ec. 16.4 S K Means Example Pick seeds Reassign clusters x ec. 16.4 S Termina*on condi*ons (K=2) x Introduc)on to Informa)on Retrieval x x Compute centroids Reassign clusters Compute centroids Reassign clusters   Several possibili*es, e.g.,   A fixed number of itera*ons.   Doc par**on unchanged.   Centroid posi*ons don t change. Converged! Does this mean that the docs in a cluster are unchanged? Introduc)on to Informa)on Retrieval Sec. 16.4 Introduc)on to Informa)on Retrieval ec. 16.4 S Lower case! Convergence Convergence of K ­Means   Why should the K ­means algorithm ever reach a fixed point?   Define goodness measure of cluster k as sum of squared distances from cluster centroid:   A state in which clusters don t change.   Gk = Σi (di – ck)2 (sum over all di i...
View Full Document

Ask a homework question - tutors are online