Data

Tools

Indexes

Applications

Experiments

Open Source Science

https://github.com/cn-tu/ids-backdoor

Contact: Maximilian Bachl, Alexander Hartl. Explores defenses against backdoors and poisoning attacks for Intrusion Detection Systems. Code for "EagerNet" is in the "eager" branch.

https://github.com/cn-tu/ids-backdoor

Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ○
    codemeta.json file
  • ○
    .zenodo.json file
  • ○
    DOI references
  • ✓
    Academic publication links
    Links to: arxiv.org, acm.org
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (4.1%) to scientific vocabulary

Keywords

adversarial-machine-learning backdoor deep-learning poisoning random-forest
Last synced: 5 months ago · JSON representation

Repository

Contact: Maximilian Bachl, Alexander Hartl. Explores defenses against backdoors and poisoning attacks for Intrusion Detection Systems. Code for "EagerNet" is in the "eager" branch.

Basic Info
Statistics
  • Stars: 8
  • Watchers: 6
  • Forks: 3
  • Open Issues: 0
  • Releases: 0
Topics
adversarial-machine-learning backdoor deep-learning poisoning random-forest
Created over 6 years ago · Last pushed over 5 years ago

https://github.com/CN-TU/ids-backdoor/blob/master/ 

# ids-backdoor
Contact: Maximilian Bachl, Alexander Hartl

This repository contains the code, the data, the plots and the tex files for our paper presented at [Big-DAMA '19](https://dl.acm.org/citation.cfm?id=3366638) ([arXiv](https://arxiv.org/abs/1909.07866)) dealing with backdoors in IDS and ways to defend against them.

Code for the *EagerNet* paper is in the *eager* branch. 

To train a random forest with the preprocessed UNSW-NB15 dataset, unzip the file `CAIA_backdoor_15.csv.gz` and run the following command to train a random forest with 100 estimators:

    ./learn.py --dataroot CAIA_backdoor_15.csv --function train --method rf --backdoor --nEstimators 100
    
To instead train a neural network, replace `--method rf` with `--method nn`. 

To reproduce Figure 6 from the paper run:

    ./learn.py --dataroot CAIA_backdoor_15.csv --method=rf --net  --function prune_backdoor --backdoor --reduceValidationSet  --nSteps 49 --pruneOnlyHarmless --depth

Owner

  • Name: CN Group, Institute of Telecommunications, TU Wien
  • Login: CN-TU
  • Kind: organization
  • Location: Vienna, Austria

Communication Networks Group, TU Wien

GitHub Events

Total
  • Watch event: 1
Last Year
  • Watch event: 1