Scribd
Upload
8 views8 pages

Major Project PhaseII

The document outlines a project focused on detecting cyber threats using machine learning techniques, including data collection, preprocessing, and model implementation. It details the system's performance metrics, achieving high accuracy and efficiency, and highlights the user-friendly Django dashboard for real-time monitoring. The project presents a cost-effective alternative to traditional SIEM tools, emphasizing its scalability and effectiveness in threat prioritization.

Uploaded by

Ziddhi Krishna
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
8 views8 pages

Major Project PhaseII

The document outlines a project focused on detecting cyber threats using machine learning techniques, including data collection, preprocessing, and model implementation. It details the system's performance metrics, achieving high accuracy and efficiency, and highlights the user-friendly Django dashboard for real-time monitoring. The project presents a cost-effective alternative to traditional SIEM tools, emphasizing its scalability and effectiveness in threat prioritization.

Uploaded by

Ziddhi Krishna
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

Cyber Threat Attack Level

Detection Using Machine


Learning
Major Project Phase-II Review
Team Members | Guide |
Department of CSE
R R Institute of Technology
Implementation (Overview)

• Data Collection: system logs, network traffic, attack datasets


• Data Preprocessing: cleaning, normalization, feature engineering
• Machine Learning Models: Random Forest, SVM, Neural Network,
Clustering
• Django Dashboard: real-time visualization & alerts
Implementation (Details)

• Severity Classification: Low, Medium, High threats


• Database Integration: SQLite (testing), PostgreSQL (production)
• Alerting System: Email, Telegram, Dashboard notifications
• Role-Based Access: Admin, Analyst, Standard User
Testing Approach

• Unit Testing: Preprocessing, ML models, alerts


• Integration Testing: Data flow between modules
• System Testing: End-to-end validation with attack simulations
• Performance Testing: 12,000 requests/sec, alerts in <3 sec
• User Acceptance Testing: Analyst & non-technical feedback
Results & Discussion (Performance
Metrics)
• Accuracy: 95.3%
• Precision: 96.1%
• Recall: 94.8%
• F1-Score: 95.4%
• Outperformed Rule-based IDS (82.5%) & DL-only models (91.7%)
Results & Discussion (System
Performance)
• Average alert response time: 2.3 seconds
• Handled 12,000 requests/second without failures
• Positive feedback: Dashboard is intuitive & severity levels clear
• Cost-effective, real-time, and scalable alternative to SIEM tools
Conclusion

• Developed ML-driven multi-attack detection system


• Severity-level classification improves incident prioritization
• Django-based dashboard enables real-time monitoring
• Achieved high accuracy, scalability, and usability
• Affordable alternative to enterprise SIEM solutions
References

• Elkouay et al. (2024) – Graph-based ML for phishing detection


• Asiri et al. (2023) – BiLSTM real-time phishing detection
• Canadian Institute for Cybersecurity – CICIDS 2017 dataset
• DARPA Intrusion Detection Evaluation dataset
• OWASP Top 10 Security Risks, NIST Cybersecurity Framework

You might also like