Top Dissertation Topics in Machine Learning, AI, and Cybersecurity

Machine Learning and AI in Cybersecurity

Optimizing Web Application Firewalls Using Ensemble Learning Techniques:
Explore how ensemble models like Random Forest and XGBoost can improve firewall accuracy and response time for modern web threats.
Deep Learning for SQL Injection and XSS Detection in Real-Time Systems:
Investigate the effectiveness of CNNs and LSTMs for SQLi and XSS detection with minimal false positives.
Transformer-based Intrusion Detection:
Evaluate the role of FT-Transformer models for securing networks using datasets like CIC-IDS and KDD Cup.
Hybrid AI Systems for DAST Attack Detection:
Develop a system combining traditional firewalls with machine learning algorithms for enhanced DAST (Dynamic Application Security Testing) attack detection.

AI-Powered Defenses Against DDoS Attacks:
Analyze AI-based algorithms to detect DDoS attacks across SDN and IoT networks effectively.
Detecting Webshells with Adversarial ML Techniques:
Explore adversarial attacks on ML models used for webshell detection and the implementation of adversarial defenses.
Anomaly-Based Detection Systems Using DNNs in Virtualized Environments:
Design deep neural network-powered anomaly detection systems for NFV infrastructures.
Enhancing DNS Security with ML and Response Policy Zones (RPZ):
Investigate how ML models integrated with RPZ can protect DNS servers from SQLi-based data exfiltration attacks.
BERT-Enhanced NLP Models for Web Request Safety:
Develop security frameworks using BERT and TextCNN to secure web and mobile applications.
AI-Driven Penetration Testing:
Explore the use of ML models for automating penetration testing frameworks that mimic real-world attack scenarios.

Mitigating Zero-Day Exploits with Hybrid Intrusion Detection Systems:
Evaluate the performance of hybrid IDS systems that combine rule-based detection with machine learning techniques.
Privacy and Ethical Implications of AI in Web Security:
Explore the trade-offs between security, privacy, and ethics in deploying AI-driven web security models.
XSS Detection Using Reinforcement Learning in Web Applications:
Develop a reinforcement learning model that adaptively detects and mitigates XSS attacks in real-time.
Lightweight ML Models for IoT Security:
Investigate how lightweight ML models can secure resource-constrained IoT devices from cyberattacks.
Real-Time Anomaly Detection in Cloud Networks:
Build scalable ML models for real-time anomaly detection to prevent disruptions in cloud environments.

AI-Powered Firewall Rule Updating for Adaptive Security:
Explore AI models that dynamically update firewall rules in response to new threats.
A Comparative Study of Word and Character N-gram Models:
Investigate the performance of word and character n-gram models in detecting web attacks within ML-powered WAFs.
Secure Coding Frameworks with ML-Powered Vulnerability Scanners:
Develop ML models that integrate with vulnerability scanners to promote secure coding practices.
Federated Learning for Distributed Intrusion Detection:
Build an IDS leveraging federated learning to analyze data across multiple sources without compromising privacy.
AI-Augmented Threat Hunting in Web Applications:
Design AI models that assist in proactive threat hunting and automated response within enterprise web applications.

421 Wandsworth Rd,Larkhall,London SW8 2RN,UK