I like this algorithm very much. I think running several steps after swapping the centers is a genius idea.
I want to ask if this repo fixed the cost computation for weighted instances. I think this may be an issue if I know the weights of some data points.
HI authors
I like this algorithm very much. I think running several steps after swapping the centers is a genius idea. I want to ask if this repo fixed the cost computation for weighted instances. I think this may be an issue if I know the weights of some data points.
Thank you for providing this nice algorithm