Implementation:Scikit learn Scikit learn EllipticEnvelope

Knowledge Sources	Scikit_learn Scikit-learn Docs
Domains	Machine Learning, Outlier Detection
Last Updated	2026-02-08 15:00 GMT

Overview

Concrete tool for detecting outliers in a Gaussian distributed dataset using a robust covariance estimate, provided by scikit-learn.

Description

The EllipticEnvelope class fits a robust covariance estimate to the data using the Minimum Covariance Determinant (MCD) algorithm and then determines which samples are outliers. It inherits from OutlierMixin and MinCovDet, combining robust covariance estimation with outlier detection capabilities.

Usage

Use this estimator when you need to detect outliers in datasets that are approximately Gaussian distributed. It is suitable for novelty detection where the training data is not contaminated, or outlier detection where a known proportion of contamination exists.

Code Reference

Source Location

• Repository: scikit-learn
• File: sklearn/covariance/_elliptic_envelope.py

Signature

class EllipticEnvelope(OutlierMixin, MinCovDet):
    def __init__(
        self,
        *,
        store_precision=True,
        assume_centered=False,
        support_fraction=None,
        contamination=0.1,
        random_state=None,
    ):

Import

from sklearn.covariance import EllipticEnvelope

I/O Contract

Inputs

Name	Type	Required	Description
store_precision	bool	No	Specify if the estimated precision is stored (default True)
assume_centered	bool	No	If True, data is assumed centered (default False)
support_fraction	float	No	Proportion of points for support of raw MCD estimate (default None)
contamination	float	No	Proportion of outliers in the data set (default 0.1, range (0, 0.5])
random_state	int, RandomState or None	No	Random number generator seed (default None)

Outputs

Name	Type	Description
location_	ndarray of shape (n_features,)	Estimated robust location
covariance_	ndarray of shape (n_features, n_features)	Estimated robust covariance matrix
precision_	ndarray of shape (n_features, n_features)	Estimated pseudo inverse matrix (if store_precision is True)

Usage Examples

Basic Usage

import numpy as np
from sklearn.covariance import EllipticEnvelope

# Generate normal data with outliers
X = np.random.randn(100, 2)
X[0] = [10, 10]  # add an outlier

envelope = EllipticEnvelope(contamination=0.1, random_state=0)
envelope.fit(X)
predictions = envelope.predict(X)
print(predictions)  # -1 for outliers, 1 for inliers

Related Pages

• Principle:Scikit_learn_Scikit_learn_Anomaly_Detection