Init k clusters depending on k
Closed this issue · 0 comments
loremacchia commented
- allocate k clusters
- get one element of the dataset
- compute distance between the point and all the other points
- choose the k-1 points with the biggest distance
- assign each point to a cluster as clustroid