Skip to content
This repository has been archived by the owner on Sep 11, 2020. It is now read-only.

Active semi-supervised clustering algorithms for scikit-learn

Notifications You must be signed in to change notification settings

datamole-ai/active-semi-supervised-clustering

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

active-semi-supervised-clustering

Active semi-supervised clustering algorithms for scikit-learn.

Algorithms

Semi-supervised clustering

  • Seeded-KMeans
  • Constrainted-KMeans
  • COP-KMeans
  • Pairwise constrained K-Means (PCK-Means)
  • Metric K-Means (MK-Means)
  • Metric pairwise constrained K-Means (MPCK-Means)

Active learning of pairwise clustering

  • Explore & Consolidate
  • Min-max
  • Normalized point-based uncertainty (NPU) method

Installation

pip install active-semi-supervised-clustering

Usage

from sklearn import datasets, metrics
from active_semi_clustering.semi_supervised.pairwise_constraints import PCKMeans
from active_semi_clustering.active.pairwise_constraints import ExampleOracle, ExploreConsolidate, MinMax
X, y = datasets.load_iris(return_X_y=True)

First, obtain some pairwise constraints from an oracle.

# TODO implement your own oracle that will, for example, query a domain expert via GUI or CLI
oracle = ExampleOracle(y, max_queries_cnt=10)

active_learner = MinMax(n_clusters=3)
active_learner.fit(X, oracle=oracle)
pairwise_constraints = active_learner.pairwise_constraints_

Then, use the constraints to do the clustering.

clusterer = PCKMeans(n_clusters=3)
clusterer.fit(X, ml=pairwise_constraints[0], cl=pairwise_constraints[1])

Evaluate the clustering using Adjusted Rand Score.

metrics.adjusted_rand_score(y, clusterer.labels_)

About

Active semi-supervised clustering algorithms for scikit-learn

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages