Unsupervised Evaluation

skfeature.utility.unsupervised_evaluation

Description

Unsupervised Evaluation provides metrics for evaluating unsupervised feature selection results, such as clustering accuracy (after best label matching) and normalized mutual information.

Usage

import numpy as np
from sklearn.datasets import load_iris

from skfeature.utility import unsupervised_evaluation

X, y = load_iris(return_X_y=True)
acc, nmi = unsupervised_evaluation.evaluation(X, n_clusters=3, y=y)

Functions

• best_map(l1, l2): permute the labels of l2 to best match l1 (Hungarian assignment)
• evaluation(X_selected, n_clusters, y): cluster the selected features and return (accuracy, normalized_mutual_information)