Enhancing Intrusion Detection Systems (IDS) with Synthetic Data
Bibliographic record
Abstract
Intrusion Detection Systems (IDSs) are fundamental to network security, diligently identifying and mitigating evolving cyber threats. However, their efficacy hinges on the availability of high-quality, diverse, and well-balanced datasets. While essential, traditional datasets such as CICIDS-2017 frequently suffer from severe class imbalances and under-representation of critical minority attacks, including brute force and infiltration. To overcome these limitations, this paper proposes integrating synthetic data generated via Variational Autoencoders (VAEs) to augment the CICIDS-2017 dataset. Our methodology involves preprocessing the original dataset to ensure data quality and generating synthetic samples that accurately replicate real-world network traffic patterns. This augmented dataset, combining real and synthetic instances, is then utilized to train state-of-the-art machine learning models, specifically Random Forest and XGBoost, for robust IDS implementation Comprehensive evaluations, employing standard classification reports and performance metrics, demonstrate that synthetic data integration significantly enhances IDS accuracy and robustness. Our results reveal substantial improvements in identifying minority-class attacks, effectively mitigating data imbalance issues, and enhancing IDS generalization capabilities. This research underscores the potential of VAE- based synthetic data generation as a powerful strategy for strengthening IDS performance.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.010 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".