Skip to content

Baukebrenninkmeijer/Bagging-and-Boosting-for-Anomaly-Detection

Repository files navigation

Natural computing final

This is the final project repository of Bauke Brenninkmeijer and Ties Robroek. The project is about using natural computing related classifiers for anomaly detection. This project is originally a GECCO challenge, found http://www.spotseven.de/gecco/gecco-challenge/gecco-challenge-2018

How to run

To get some sense of the project is quite easy. The different detectors we created are in the detectors directory. To get immediate results, the EvaluationMain.R file can be run. This runs all the detectors on the train data with a 2/3 train 1/3 test split. The results will be given in true positive, false positive, true negative and false negative, as well as the corresponding F1-score. Additionally, the boostingEnsemble.R file can be run for a 10-fold cross validation score of Adaboost, XGBoost and an ensemble of them.

About

This project compares multiple bagging and boosting methods for anomaly detection for the Gecco challenge.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages