I want to use a clustering algorithm for a data set that each row/record has 3 features/columns. All the features are integers.
First, I have tried DBSCAN method because it is a method that I know well. I have used the NearestNeighbors algorithm to choose eps and minpts values. The problem is that even if I choose a big N value, the N neighbors for many records are 0, because many records are the same. I don't want to delete these duplicate records.
Is there a particular clustering method for this scenario?