Anomaly Detection comparison

Simple demonstration of Anomaly Detection methods implemented in emlearn

# Example adapted from
# "Comparing anomaly detection algorithms for outlier detection on toy datasets"
# Author: Alexandre Gramfort <alexandre.gramfort@inria.fr>
#         Albert Thomas <albert.thomas@telecom-paristech.fr>
# License: BSD 3 clause
# https://scikit-learn.org/stable/auto_examples/miscellaneous/plot_anomaly_comparison.html#sphx-glr-auto-examples-miscellaneous-plot-anomaly-comparison-py

import os.path

import emlearn
import numpy
import pandas
import seaborn
import matplotlib.pyplot as plt

#matplotlib.rcParams["contour.negative_linestyle"] = "solid"


try:
    # When executed as regular .py script
    here = os.path.dirname(__file__)
except NameError:
    # When executed as Jupyter notebook / Sphinx Gallery
    here = os.getcwd()

Create dataset

Using a simple multi-class dataset included with scikit-learn

def make_datasets(n_samples = 300, outliers = 0.15, seed=42):
    from sklearn.datasets import make_moons, make_blobs

    rng = numpy.random.RandomState(seed)

    n_outliers = int(outliers * n_samples)
    n_inliers = n_samples - n_outliers

    # Define datasets
    blobs_params = dict(random_state=0, n_samples=n_inliers, n_features=2)
    datasets = [
        make_blobs(centers=[[0, 0], [0, 0]], cluster_std=0.5, **blobs_params)[0],
        make_blobs(centers=[[2, 2], [-2, -2]], cluster_std=[0.5, 0.5], **blobs_params)[0],
        make_blobs(centers=[[2, 2], [-2, -2]], cluster_std=[1.5, 0.3], **blobs_params)[0],
        4.0
        * (
            make_moons(n_samples=n_samples, noise=0.05, random_state=0)[0]
            - numpy.array([0.5, 0.25])
        ),
        #14.0 * (numpy.random.RandomState(42).rand(n_samples, 2) - 0.5),
    ]

    def add_outliers(X):
        return numpy.concatenate([X, rng.uniform(low=-6, high=6, size=(n_outliers, 2))], axis=0)
    datasets = list(map(add_outliers, datasets))

    return datasets

outliers_fraction = 0.15
datasets = make_datasets(outliers=outliers_fraction)

Models to compare

from sklearn.covariance import EllipticEnvelope
from sklearn.ensemble import IsolationForest
from sklearn.mixture import GaussianMixture, BayesianGaussianMixture

anomaly_algorithms = [
    ("Elliptic Envelope", EllipticEnvelope(contamination=outliers_fraction)),
    ("GMM (2, full)", GaussianMixture(n_components=2, covariance_type='full')),
    ("GMM (4, full)", GaussianMixture(n_components=4, covariance_type='full')),
    #("Gaussian Mixture model (32, full)", GaussianMixture(n_components=4, covariance_type='diag', random_state=1)),
    ("Baysian GMM ", BayesianGaussianMixture(n_components=12,
            covariance_type='diag', random_state=1, n_init=4,
            degrees_of_freedom_prior=1.1, max_iter=20)
    ),

    # Not yet supported
    #( "Isolation Forest", IsolationForest(contamination=outliers_fraction, random_state=42)),
    #("One-Class SVM", svm.OneClassSVM(nu=outliers_fraction, kernel="rbf", gamma=0.1)),
]

Plotting tools

Plots the anomaly score landscape

def plot_results(ax, model, X):
    res = 20
    xx, yy = numpy.meshgrid(numpy.linspace(-7, 7, res), numpy.linspace(-7, 7, res))

    try:
        y_pred = model.score_samples(X)
        Z = model.score_samples(numpy.c_[xx.ravel(), yy.ravel()])
    except FloatingPointError as e:
        print(e)
        return

    # Normalize the anomaly scores to 0.0 -> 1.0
    from sklearn.preprocessing import MinMaxScaler
    scaler = MinMaxScaler()
    def nomalize_score(s, fit=False):
        s = -1.0 * s.reshape(-1, 1)
        if fit:
            scaler.fit(s)
        return scaler.transform(s)[:,0]

    Z = nomalize_score(Z, fit=True)
    y_pred = nomalize_score(y_pred, fit=True)

    cmap = seaborn.color_palette("rocket", as_cmap=True)

    Z = Z.reshape(xx.shape)
    ax.contour(xx, yy, Z, levels=numpy.linspace(0.0, 1.0, 5), linewidths=0.6, colors="black")
    ax.contourf(xx, yy, Z, cmap=cmap)

    # Plot datapoints
    seaborn.scatterplot(ax=ax, x=X[:, 0], y=X[:, 1], s=10, hue=y_pred, palette=cmap, legend=False)

    ax.set_xlim(-7, 7)
    ax.set_ylim(-7, 7)
    ax.set_xticks(())
    ax.set_yticks(())

Run comparison

fig, axs = plt.subplots(
            ncols=len(anomaly_algorithms),
            nrows=len(datasets),
            figsize=(len(anomaly_algorithms) * 2 + 4, 12.5),
            sharex=True, sharey=True,
)

plt.subplots_adjust(left=0.02, right=0.98, bottom=0.001, top=0.96, wspace=0.05, hspace=0.01)
for i_dataset, X in enumerate(datasets):
    for i_algorithm, (name, model) in enumerate(anomaly_algorithms):

        # Train model
        print(f"Trying {name}")
        try:
            model.fit(X)
        except FloatingPointError as e:
            print(e)
            continue

        # Convert to C
        cmodel = emlearn.convert(model, method='inline')

        # Visualize output
        ax = axs[i_dataset, i_algorithm]
        plot_results(ax, cmodel, X)

        if i_dataset == 0:
            ax.set_title(name, size=18)

plt.show()

Trying Elliptic Envelope
Trying GMM (2, full)
Trying GMM (4, full)
Trying Baysian GMM
/home/docs/checkouts/readthedocs.org/user_builds/emlearn/envs/latest/lib/python3.8/site-packages/sklearn/mixture/_base.py:268: ConvergenceWarning: Initialization 4 did not converge. Try different init parameters, or increase max_iter, tol or check for degenerate data.
  warnings.warn(
Trying Elliptic Envelope
Trying GMM (2, full)
Trying GMM (4, full)
Trying Baysian GMM
/home/docs/checkouts/readthedocs.org/user_builds/emlearn/envs/latest/lib/python3.8/site-packages/sklearn/mixture/_base.py:268: ConvergenceWarning: Initialization 4 did not converge. Try different init parameters, or increase max_iter, tol or check for degenerate data.
  warnings.warn(
Trying Elliptic Envelope
Trying GMM (2, full)
Trying GMM (4, full)
Trying Baysian GMM
/home/docs/checkouts/readthedocs.org/user_builds/emlearn/envs/latest/lib/python3.8/site-packages/sklearn/mixture/_base.py:268: ConvergenceWarning: Initialization 4 did not converge. Try different init parameters, or increase max_iter, tol or check for degenerate data.
  warnings.warn(
Trying Elliptic Envelope
Trying GMM (2, full)
Trying GMM (4, full)
underflow encountered in exp
Trying Baysian GMM
/home/docs/checkouts/readthedocs.org/user_builds/emlearn/envs/latest/lib/python3.8/site-packages/sklearn/mixture/_base.py:268: ConvergenceWarning: Initialization 4 did not converge. Try different init parameters, or increase max_iter, tol or check for degenerate data.
  warnings.warn(