k-means clustering is an algorithm that searches for a given numebr of clusters within some unlabeled multidimensional dataset. The optimal clustering is found as follow:
The “cluster centre” is the artihmetic mean of all of the points belonging to the cluster.
Each repeatition will always result in an better estimate of the cluster characteristics.
The globally optimal result may not be achieved: As is the cause with all optimisations, although the result is improving,
arm61
commented
3 years ago
At https://mccluskey.scot/trad_ml_methods/k_means.html
k-means clustering is an algorithm that searches for a given numebr of clusters within some unlabeled multidimensional dataset. The optimal clustering is found as follow:
Each repeatition will always result in an better estimate of the cluster characteristics.
The globally optimal result may not be achieved: As is the cause with all optimisations, although the result is improving,