Unsupervised Anomaly Detection

Motivation

A Notebook where I implement differents Unsupervised anomaly detection algorithms on a simple exemple. The goal was to understand how the different algorithms works and their differents caracteristics. I have created this notebook after reading this article: https://iwringer.wordpress.com/2015/11/17/anomaly-detection-concepts-and-techniques/ I implement the algorithms cited in the article and add some.

Algorithm implemented :

• Cluster based anomaly detection (K-mean)
• Repartition of data into categories then Gaussian/Elliptic Enveloppe on each categories separately
• Markov Chain
• Isolation Forest
• One class SVM
• RNN (comparison between prediction and reality)