AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

Interpreting Unsupervised Anomaly Detection in Security via Rule Extraction

Ruoyu Li, Qing Li, Yu Zhang, Dan Zhao, Yong Jiang, Yong Yang

NeurIPS 2023 · Label-efficient and robust learning › OOD and anomaly detection · approach: Anomaly detection with contaminated data · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace IC-Tree with K-Means (k=10) (ablation)IC-Tree + CBE · CIC-IDS2017fidelity: 0.9856 -> 0.9731Hurt
replace IC-Tree with K-Means (k=5) (ablation)IC-Tree + CBE · CIC-IDS2017fidelity: 0.9856 -> 0.9735Hurt
use only CBE without IC-Tree (ablation)IC-Tree + CBE · CIC-IDS2017fidelity: 0.9856 -> 0.9735Hurt
use hypercubes as rules instead of CBE (ablation)IC-Tree + CBE · CIC-IDS2017fidelity: 0.9856 -> 0.9652Hurt
use hypercubes as rules instead of CBE (ablation)IC-Tree + CBE · TON-IoTfidelity: 0.9840 -> 0.0647Hurt
replace IC-Tree with K-Means (k=10) (ablation)IC-Tree + CBE · TON-IoTfidelity: 0.9840 -> 0.9802Hurt
use only CBE without IC-Tree (ablation)IC-Tree + CBE · TON-IoTfidelity: 0.9840 -> 0.9784Hurt
add 10% random noise to training dataOurs on iForest · random noisefidelity: 0.9876 -> 0.5148Hurt
add 10% random noise to training dataOurs on AE · random noisefidelity: 0.9829 -> 0.9739No clear effect
add 10% mislabeled attack data to trainingOurs on AE · mislabeled noisefidelity: 0.9997 -> 0.9987No clear effect
use rule extraction Ours instead of EGDTrule extraction · CIC-IDS2017 AEfidelity: 0.533 -> 0.9835Improved
use Ours instead of UAD rule extractionrule extraction · TON-IoT AEfidelity: 0.1499 -> 0.9996Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy