Research objectives

• Critically evaluate existing SI and machine learning approaches for malware detection and prevention, identifying limitations in predictive, cognitive, and autonomous threat intelligence.
• Build a comprehensive dataset integrating historical malware, ransomware, polymorphic payloads, behavioural indicators, and community-submitted emerging samples.
• Design a hybrid SI architecture combining deep learning for pattern recognition, reinforcement learning for agentic decision-making, and cognitive reasoning layers for inferring novel behaviours from incomplete signals.
• Develop and implement automated malware simulation and predictive defence for real-time analysis and mitigation.
• Evaluate effectiveness against simulated zero-day and polymorphic malware, benchmarking against traditional reactive security models.

• Objective 1 (critical evaluation): the literature review and bibliography, informing all five research questions.
• Objective 2 (integrated dataset): methodology phase 1, and the basis for RQ-005 on community submissions.
• Objective 3 (hybrid architecture): phase 2, addressing RQ-001 to RQ-003.
• Objective 4 (simulation and predictive defence): phase 3, addressing RQ-004 on governed mitigation.
• Objective 5 (evaluation): phase 4, answering RQ-001 and RQ-004 against reactive baselines.

Each objective is tracked through experiments and datasets in the research workspace. Public pages report progress as it is evidenced; nothing is marked complete until the supporting results are recorded with provenance.