Awesome Anomaly Detection

A list of Papers on anomaly detection. You are welcome to open an issue and pull your requests if you think any paper that is important but not are inclueded in this repo. The papers are orgnized in classical method, deep learning method, application and survey.

Classical Method

• Isolation Forest - ICDM 2008.

• LOF: Identifying Density-Based Local Outliers - SIGMOD 2000.

• Extended Isolation Forest

• Support Vector Method for Novelty Detection - NIPS 2000

One-Class Classification

• One-Class SVMs for Document Classification - JMLR 2001.

• Support Vector Data Description

• Can I Trust My One-Class Classification?

• Efficient Anomaly Detection via Matrix Sketching - NIPS 2018

PCA-based

• robust deep and inductive anomaly detection - ECML PKDD 2017

• A loss framework for calibrated anomaly detection - NIPS 2018

Clustering

• A Practical Algorithm for Distributed Clustering and Outlier Detection - NIPS 2018

Correlation

• Detecting Multiple Periods and Periodic Patterns in Event Time Sequences - CIKM 2017.

Ranking

• ranking causal anomalies via temporal and dynamical analysis on vanishing correlations - KDD 2016.

Deep Learning Method

Generative Methods

• Variational Autoencoder based Anomaly Detection using Reconstruction Probability

Auto-encoder

• Learning sparse representation with variational auto-encoder for anomaly detection

• Anomaly Detection with Robust Deep Autoencoders - KDD 2017.

• DEEP AUTOENCODING GAUSSIAN MIXTURE MODEL FOR UNSUPERVISED ANOMALY DETECTION - ICLR 2018.

• Generative Probabilistic Novelty Detection with Adversarial Autoencoders - NIPS 2018

Variational Auto-encoder

• Multidimensional Time Series Anomaly Detection: A GRU-based Gaussian Mixture Variational Autoencoder Approach - ACML 2018

• A Multimodel Anomaly Detector for Robot-Assisted Feeding Using an LSTM-based Variational Autoencoder - IEEE Robotics and Automation Letters 2018.

GAN based

• Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery - IPMI 2017.

• Efficient-GAN-Based Anomaly Detection ICLR Workshop 2018.

• Anomaly detection with generative adversarial networks - Reject by ICLR 2018, but was used as baseline method in recent published NIPS paper.

Hypersphereical Learning

• Anomaly Detection in Dynamic Networks using Multi-view Time-Series Hypersphere Learning - CIKM 2017.

• Deep into Hypersphere: Robust and Unsupervised Anomaly Discovery in Dynamic Networks - IJCAI 2018.

One-Class Classification

• High-dimensional and large-scale anomaly detection using a linear one-class SVM with deep learning - Pattern Recognition 2018.

• Optimal single-class classification strategies - NIPS 2007

• Deep One-Class Classification - ICML 2018.

Energy-based

• Deep structured energy based models for anomaly detection - ICML 2016

Time series

• A Generalized Student-t Based Approach to Mixed-Type Anomaly Detection - AAAI 2013

• Stochastic Online Anomaly Analysis for Streaming Time Series - IJCAI 2017

• Long short term memory networks for anmomaly detection in time series

• LSTM-based Encoder-Decoder for Multi-sensor Anomaly Detection - ICML 2016 Workshop.

Interpretation

• Contextual Outlier Interpretation -IJCAI 2018

Evaulation Metrics

• Precision and Recall for Time Series - NIPS 2018

Geometric transformation

• Deep Anomaly Detection Using Geometric Transformations - NIPS 2018

FeedBack

• Incorporating Feedback into Tree-based Anomaly Detection - KDD 2017 Workshop on Interactive Data Exploration and Analytics.

• Feedback-Guided Anomaly Discovery via Online Optimization - KDD 2018.

Anomaly Detection Applications

KPI

• Unsupervised Anomaly Detection via Variational Auto-Encoder for Seasonal KPIs in Web Applications - WWW 2018.

Log

• DeepLog: Anomaly Detection and Diagnosis from System Logs through Deep Learning - CCS 2017.

• Mining Invariants from Logs for System Problem Detection - USENIX 2010

Survey

• Anomaly detection in dynamic networks: a survey

• Anomaly Detection : A Survey

• A Survey of Recent Trends in One Class Classification

• A survey on unsupervised outlier detection in high‐dimensional numerical data