Do k-means training algorithm
Do k-means training algorithm
sample point
k centers
max training iter limitation
Get initialized k center points
Get initialized k center points
sample point
k centers
Happened if K has changed while iteration
Print loss and step information to console, override it to output to another place
Print loss and step information to console, override it to output to another place
current step
loss value: sum distance for each point to it's center
the count of the point
(Since version ) see corresponding Javadoc for more information.
Implementation of K-Means++ on geographical points.
type of point
paper: k-means++: The Advantages of Careful Seeding