Enabling Cross-Context Learning with Knowledge Graphs for Anomaly Detection

Experiments to perform anomaly detection using autoencoder-based techniques. Experiments make use of cross-context learning to support dynamic adaptation of anomaly detectors to new deployment contexts.

Running

1. Clone this project: git clone....
2. Preferably create a virtual environment (conda create --name crosscontext) and activate it (conda activate crosscontext ).
3. cd to the project's root folder and install all required packages: pip install -r requirements.txt.
4. experiments.py contains cross-context experiments for various transfer criteria.