clustering

The Turi Create clustering toolkit provides tools for unsupervised learning tasks, where the aim is to consolidate unlabeled data points into groups. Points that are similar to each other should be assigned to the same group and points that are different should be assigned to different groups.

The clustering toolkit contains two models: K-Means and DBSCAN.

Please see the documentation for each of these models for more details, as well as the clustering chapter of the User Guide.

dbscan

dbscan.create	Create a DBSCAN clustering model.
dbscan.DBSCANModel	DBSCAN clustering model.

kmeans

kmeans.create	Create a k-means clustering model.
kmeans.KmeansModel	K-means clustering model.