Dear researchers
I have one question regarding to clustering by k-means.
I have one series of data that I divided them into 2 clusters with k-means.
in the first clusters there are 2 points that I want to remove them since they are not that much similar to the first cluster and these two points are more similar to the second cluster.
My professor told me to calculate the likelihood ratio of the points rather than the first and second cluster.
then divide these ratio and determine a threshold.
Do you have any idea how to calculate it in MATLAB?
thanks for your time.