Hybrid ML SIEM Paper

I need an 8- to 10-page conference paper that presents a hybrid machine-learning Security Information and Event Management (SIEM) framework combining Random Forest and Isolation Forest for network-threat detection. The manuscript must follow either Springer LNCS or Scopus proceedings guidelines, complete with the correct template, figure sizing, and reference style.

Core structure
• Introduction and literature review that positions the problem, surveys recent SIEM advances, and justifies the hybrid approach.
• Methodology and data analysis describing data-preprocessing, feature engineering, model building in scikit-learn, and experimental evaluation on publicly available cybersecurity datasets (e.g., CIC-IDS 2017, UNSW-NB15, or similar).
• Conclusion and future work highlighting detection accuracy, false-positive reduction, and directions for real-time deployment.

Technical requirements
• 6–7 clearly numbered mathematical formulas (e.g., precision, recall, F1, G-mean, ensemble weighting) set with the template’s equation environment.
• A few well-labelled diagrams: system architecture, data-flow, and comparative ROC/PR curves.
• All in-text citations and reference list strictly in APA style.
• Implementation notes reference scikit-learn (RandomForestClassifier, IsolationForest) and any supporting Python tools such as pandas, NumPy, and Matplotlib.

Deliverables
1. Editable source files (Word / LaTeX plus figures).
2. A compiled PDF ready for direct submission.
3. A short README outlining dataset links, Python version, and command line to reproduce results.

Acceptance criteria
• Conforms to Springer or Scopus template without formatting warnings.
• Plagiarism-free, originality score ≤ 5 %.
• Experiments reproducible with the included notebook or script.

Deadline: 07 March 2026. Please keep periodic checkpoints so I can review drafts, figures, and the reference list along the way.