Please provide your feedback in this short Flings' survey.
fling logo of Resource-Efficient Supervised Anomaly Detection Classifier

Resource-Efficient Supervised Anomaly Detection Classifier

version 1.0.0 — June 02, 2021

Contributors 2

View All

Summary

Resource-Efficient Supervised Anomaly Detection Classifier is a scikit-learn classifier for resource-efficient anomaly detection that augments either Random-Forest or XGBoost classifiers to perform well in a resource-constrained setting. To that end, Resource-Efficient Supervised Anomaly Detection Classifier offers reduced memory footprint and computation as compared to Random-Forest and XGBoost.

The key idea behind Resource-Efficient Supervised Anomaly Detection Classifier is first to train a small model that is sufficient to correctly classify the majority of the queries. Then, using only subsets of the training data, train expert models for these fewer harder cases where the small model is at high risk of making a classification mistake.

We are happy to help you integrating RADE into your use case.
If you would like our help, please contact us:
Yaniv Ben-Itzhak, ybenitzhak@vmware.com
Shay Vargaftik, shayv@vmware.com

Requirements

Prerequisites:

  • CMake 3.13 or higher
  • numpy
  • pandas
  • sklearn
  • xgboost

Instructions

Prerequisities installation:

- Install CMake 3.13 or higher.

- The python packages can be installed manually or by running:

   pip3 install -r requirments.txt

 

Usage:

- This classifier is implemented as a sci-kit classifier.

- example_program.py contains an example that compares RADE to Random Forest and XGBoost.

- Another example code for using RADE is: 

    from rade_classifier import RadeClassifier

    from sklearn.datasets import make_classification

    from sklearn.metrics import classification_report

    from sklearn.model_selection import train_test_split



    X, y = make_classification(n_samples=1000000, n_features=4,

                                n_informative=2, n_redundant=0,

                                random_state=0, shuffle=False, weights=[0.99, 0.01])



    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, stratify=y, random_state=42)



    clf = RadeClassifier()



    clf.fit(X_train, y_train)

    y_predicted = clf.predict(X_test)

    print(classification_report(y_test, y_predicted, digits=5))

Changelog

Version 1.0.0

  • Added OSL file

Similar Flings

No similar flings found. Check these out instead...
Aug 17, 2020
fling logo of Infrastructure Deployer for vCloud NFV

Infrastructure Deployer for vCloud NFV

version 3.3

Infrastructure Deployer for vCloud NFV is an automation-based deployment tool used for setting up the VMware vCloud NFV platform (NFV 3.2 VCD edition).

Nov 09, 2022
UPDATED
fling logo of Workspace ONE ConQuest

Workspace ONE ConQuest

version 2

ConQuest configures Meta Quest 2 VR consumer devices for enrollment into Workspace ONE UEM. ConQuest is an application for Windows that can be run once a Meta Quest 2 device is set to developer mode and connected to the Windows device via USB. (Mac/Linux versions to follow)

Feb 17, 2011
fling logo of vCenter XVP Manager and Converter

vCenter XVP Manager and Converter

version 1.0

XVP Manager provides non-vSphere platforms centralized control.

Nov 26, 2019
fling logo of Kubewise

Kubewise

version 1.1.0

Kubewise is a simple multi-platform desktop client for Kubernetes®.

Mar 07, 2010
fling logo of Weasel

Weasel

version 4.1

Weasel is an Operating System installer. When you insert the ESX Installation DVD, this program guides you through the steps of network configuration, disk selection, etc. Or it can perform an automated install based on a script similar to Redhat kickstart scripts.

Jul 19, 2022
fling logo of Easy Deploy for NSX Advanced Load Balancing

Easy Deploy for NSX Advanced Load Balancing

version 3.1.6

Easy Deploy for NSX Advanced Load Balancer (formerly Avi Networks) Fling is a virtual appliance that helps you deploy Avi in a handful of clicks!

View More