Magyarázható és megbízható többügynökös generatív mesterséges intelligencia a döntéshozatal támogatásához

Debreceni Egyetem

Informatikai Tudományok Doktori Iskola

Bérczes Tamás

informatikai tudományok

The PhD topic focuses on developing explainable and reliable multi-agent generative AI systems for decision support applications. Multiple autonomous AI agents collaborate using generative models (e.g., large language models - LLMs, generative adversarial networks - GANs) to aid complex decisions, ensuring transparency through techniques such as attention mechanisms, feature attribution methods, and post-hoc explanations integration. Reliability is emphasized via robustness, fault tolerance, data privacy, and bias mitigation, including the application of security protocols in critical settings. The research involves building prototypes for real-world domains, such as healthcare (e.g., diagnostic decisions), finance (e.g., risk assessment), or logistics (e.g., supply chain optimization), with empirical evaluations assessing system performance using explainability metrics (e.g., faithfulness, plausibility) and reliability indicators (e.g., accuracy under uncertainty, robustness to adversarial attacks). Expected outcome: An innovative framework that enhances the applicability of generative AI in critical decision environments, fostering human-AI collaboration. The topic is interdisciplinary, combining AI, machine learning, ethical considerations, and human-AI interaction, ideal for motivated candidates with strong programming and mathematical backgrounds.



References
• Liu, X., et al. (2025). Explainable AI in Multi-Agent Systems: Advancing Transparency with Layered Prompting. ResearchGate.
• Smith, J., & Johnson, A. (2025). Generative AI for decision-making: A multidisciplinary perspective. Journal of Innovation in Digital Ecosystems.
• Calvaresi, D., et al. (Eds.). (2025). Explainable, Trustworthy, and Responsible AI and Multi-Agent Systems. Springer.
• Kim, H., et al. (2025). Multi-task reinforcement learning and explainable AI-Driven platform for clinical decision support. Scientific Reports.
• Wang, L., et al. (2025). Designing with Multi-Agent Generative AI: Insights from Industry Practitioners. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies.

ITDI

2026-01-15