LapScore¶

skfeature.function.similarity_based.lap_score

Description¶

Laplacian Score is an unsupervised method. It builds an affinity graph over the samples and scores each feature by how well it preserves the local manifold structure of the data; lower scores indicate better features.

Usage¶

import numpy as np
from sklearn.datasets import load_iris

from skfeature.function.similarity_based import lap_score

X, y = load_iris(return_X_y=True)  # labels are optional for this unsupervised method

# get a score for every feature (aligned with SelectKBest)
score = lap_score.lap_score(X, y)

# or get the indices of the selected features
selected = lap_score.lap_score(X, y, mode="index")

Parameters¶

X: numpy array, shape (n_samples, n_features) — input data
y: numpy array, shape (n_samples,) or None — optional class labels (unsupervised)
**kwargs: optional W sparse affinity matrix, shape (n_samples, n_samples)
mode: {{"rank", "index"}}, default "rank" — "rank" returns an array of feature indices ordered by importance and aligned with sklearn.feature_selection.SelectKBest; "index" returns the indices of the selected features with the most important one first

Returns¶

score: numpy array, shape (n_features,) — ranking score of every feature, aligned with sklearn.feature_selection.SelectKBest

References¶

Original implementation from the DMML Lab@ASU Feature Selection Repository.