AI
A Multi Method Importance and Performance Efficiency Analysis of Topological Metrics for Natural Visibility Graph Based Cyber Attack Detection
Publication: cs.AI updates on arXiv.org
Author: Ali Melih Kanca, Ilker Turker
Source published: 05/10/2026, 04:00:00 UTC
Briefing created: 05/10/2026, 04:27:04 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 05/10/2026, 04:03:24 UTC.
Briefings
A study evaluated 21 Natural Visibility Graph (NVG) topological metrics using 21 distinct methods (SHAP, Permutation, Boruta, RFE) on the CICIDS2018 dataset. The consensus ranking identified Top3 as the most effective configuration, achieving 97.148% accuracy and 97.055% weighted F1 score, while reducing runtime from 14,961.39 seconds to 589.22 seconds. The three highest ranked metrics were avg_clustering_coeff_median, avg_clustering_coeff_std, and avg_clustering_coeff_mean.
Key points
- The study evaluated 21 NVG derived topological metrics using 21 distinct importance analysis methods (SHAP, Permutation, Boruta, RFE) on the CICIDS2018 dataset.
- Top3 achieved the highest observed mean performance with 97.148% accuracy and 97.055% weighted F1 score, compared to Full21's 95.999% accuracy.
Context and limitations
Results are based on the CICIDS2018 dataset and stratified 5-fold cross validation; performance metrics are observed mean values, not generalizable to unseen data.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 05/10/2026, 04:27:04 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeidit7jay6x7gzbp2bvbv5ntbw72dtnvbcphzaw4ihbpifog4urjfu