Score Clusters by KNN Lookup

By for October 24, 2016

152 views
143 downloads

Language

Report Abuse
Assign new data to clusters using K Nearest Neighbors lookup against existing cluster assignments
Score Clusters by KNN Lookup takes existing data with cluster assignments, for example as produced by Fit DBSCAN Clustering or Fit Hierarchical Clustering. The module also takes a new, unlabeled dataset, and assigns the new data into clusters by using a K Nearest Neighbors lookup against the existing cluster assignments. The module uses R knn function to perform the lookup. The module code is available [here][1]. [1]: https://github.com/Azure/custmod/tree/master/Clusterers