Code for conducting experiments on article "Improved Anomaly Detection by Using the Attention-Based Isolation Forest"
$ pip install -r requirements.txt
To run the code, the hydra package is used to set configurations:
There are two modes for running the code, which are set in the config.yaml:
- optimization
- inference
mode: "optimization"
defaults:
- inference
- optimization
The corresponding configuration files are used to set up datasets, model parameters (isolation forest or attention based isolation forest and other parametrs) and logs.