MALTESQUE2020-IaC-Novelty-Detection
Replication package for the paper "Singling the Odd Ones Out: A Novelty Detection Approach to Find Defects in Infrastructure-as-Code".
The dataset and the Machine-Learning scripts for Novelty Detection can be found on Kaggle.
The original dataset can be found on Zenodo.
Repository structure
Repository
|-data/
|- isoforest.json: output for the IsolationForest model validation.
|- localoutlier.json: output for the LocalOutlierFactor model validation.
|- ocsvm.json: output for the OneClassSVM model validation.
|- rf.json: output for the RandomForest model validation.
|-plots/
|- auc.png: boxplots of the four techniques' AUC-PR.
|- mcc.png: boxplots of the four techniques' MCC.
|-plots.py: script to generate plots.