- BaggingRandomMiner is an ensemble of weak one-class classifiers based on dissimilarities:
- BRM can work with only numerical features. If you have nominal features, a solution is to use OneHotEncoder. An example of using BRM with nominal features can be found HERE.
- The following link includes a working example in Google Colab: https://colab.research.google.com/drive/1Qsi1UdUKBpBdL3bJqiaoXMgjYzhBJdVl?usp=sharing
- You can install BRM as a python package by using pip install brminer.
- More information about training and classification stages can be seen HERE
- A practical example about how to use BRM can be seen in BRM/tree/master/Example.
- BRM has been used in several papers, such as:
- A one-class classification approach for bot detection on Twitter
- Pattern-Based and Visual Analytics for Visitor Analysis on Websites
- One-Class Classification in Images and Videos Using a Convolutional Autoencoder With Compact Embedding
- HMMs based masquerade detection for network security on with parallel computing
- Improving the Dense Trajectories Approach Towards Efficient Recognition of Simple Human Activities
- Big Data Analytics in Cyber Security: Network Traffic and Attacks
- m-OCKRA: An Efficient One-Class Classifier for Personal Risk Detection, Based on Weighted Selection of Attributes
- Obfuscation of Malicious Behaviors for Thwarting Masquerade Detection Systems Based on Locality Features
For more information about BRM, please visit: J. Benito Camiña, M.A. Medina-Pérez, R. Monroy, O. Loyola-González, L. A. Pereyra-Villanueva, L. C. González-Gurrola "Bagging-RandomMiner: A one-class classifier for file access-based masquerade detection," Machine Vision and Applications, vol. 30, no. 5, pp. 959-974, 2019.