Development of a Malicious Network Traffic Intrusion Detection System Using Deep Learning
Bibliographic record
Abstract
With the exponential surge in the number of internet-connected devices, the attack surface for potential cyber threats has correspondingly expanded. Such a landscape necessitates the evolution of intrusion detection systems to counter the increasingly sophisticated mechanisms employed by cyber attackers. Traditional machine learning methods, coupled with existing deep learning implementations, are observed to exhibit limited proficiency due to their reliance on outdated datasets. Their performance is further compromised by elevated false positive rates, decreased detection rates, and an inability to efficiently detect novel attacks. In an attempt to address these challenges, this study proposes a deep learning-based system specifically designed for the detection of malicious network traffic. Three distinct deep learning models were employed: Deep Neural Networks (DNN), Long Short-Term Memory (LSTM), and Gated Recurrent Units (GRU). These models were trained using two contemporary benchmark intrusion detection datasets: the CICIDS 2017 and the Coburg Intrusion Detection Data Sets (CIDDS). A robust preprocessing procedure was conducted to merge these datasets based on common and essential features, creating a comprehensive dataset for model training. Two separate experimental setups were utilized to configure these models. Among the three models, the LSTM displayed superior performance in both experimental configurations. It achieved an accuracy of 98.09%, a precision of 98.14%, an F1-Score of 98.09%, a True Positive Rate (TPR) of 98.05%, a True Negative Rate (TNR) of 99.69%, a False Positive Rate (FPR) of 0.31%, and a False Negative Rate (FNR) of 1.95%.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".