python - Scipy hierarchial clustering - clusterize new vector -
i clusterizing bunch of feature vectors using scipy linkage
ward
method.
i want predictive model works in 2 steps:
- training data clusterized
- a new vector comes, distance between vector , each cluster computed, new vector assigned "nearest" cluster's label
how can compute distances between new vector , precomputed clusters?
use euclidean metric? think precomputed clusters have center. can calculate distanc between every center , new vector.
Comments
Post a Comment