After clustering on a subset of time series, how can I associate the remaining time series with already created clusters?

Question

I would like to know if there is a way to associate time series with existing clusters?

In practice, I considered a subset of time series and for each I extracted some features (after which I applied the k-means and grouped similar ones) having 6 clusters.

Is it possible to insert the remaining time series directly into one of the clusters already created in which there are similar time series?

Answer 1

You can get the centroid of each cluster, compute the distance of the new time-serie to each centroid and insert that time-serie to the nearest cluster.