Clustering 5: K-means objective and convergence
38.6 هزار بار بازدید -
11 سال پیش
-
Full lecture:
Full lecture: http://bit.ly/K-means
K-means algorithm attempts to minimize the intra-cluster variance (aggregate distance from the cluster centroid to the instances in the cluster). K-means converges to a local minimum, so different initializations will result in different clusterings. K-means does not guarantee that similar (nearby) instances will end up in the same cluster.
K-means algorithm attempts to minimize the intra-cluster variance (aggregate distance from the cluster centroid to the instances in the cluster). K-means converges to a local minimum, so different initializations will result in different clusterings. K-means does not guarantee that similar (nearby) instances will end up in the same cluster.
11 سال پیش
در تاریخ 1392/10/29 منتشر شده
است.
38,637
بـار بازدید شده