The aim of IAT is to research and develop new methods, techniques and software solutions for Intelligent Agents and Multi-Agent Systems and their applications in complex real-world problems, focusing on autonomous decision-making, coordination and adaptation mechanisms for building and managing open large-scale distributed systems.
An intelligent multi-agent system comprises a number of intelligent software agents that can autonomously function and interact in a complex dynamic environment. The agents are intelligent software entities that have the abilities to perceive the environment, reason about own and other agents' goals, strategies and actions, and make decisions about the best course of actions in order to achieve desired outcomes. Each agent is autonomous as it controls its internal states, decision-making processes, behaviour and interactions. The agents can adapt to changes in the environment, behaviour of other agents and learn from experience. They interact with other agents through direct communication or indirectly through their actions on the environment. The overall behaviour and performance of the system depends on the individual agents' behaviours and interactions that can often result in an emergent behaviour of the system.
Our research in Intelligent Agents and Multi Agent Systems focuses on autonomous decision-making, coordination and adaptation mechanisms for agent systems situated in open dynamic environments characterised by the presence of changing, incomplete and uncertain information. The research areas include complex agent negotiations and collective decision-making, distributed learning and adaptation in multi-agent systems, dynamic interactions and organisational mechanisms, and intelligent mobile agents and service-oriented agents. The application areas involve smart infrastructure (smart energy grids, smart transport), adaptive SLA and QoS management in cloud and service-oriented environments, collaborative e-business and virtual organisations, and complex adaptive systems.
Complex agent negotiations and decision-making
Distributed learning and adaptation in multi-agent systems
- Preference elicitation, aggregation and evaluation
- Design, convergence, optimality of negotiation strategies
- Multi-attribute, multi-lateral, multi-stage negotiation mechanisms
- Heterogenous, collective and emergent decision-making
- Asynchronous multi-party decision-making and negotiation
- Qualitative, fuzzy, linguistic and perception-based negotiations
Interactions and dynamic organisations
- Learning and evolving decision models and strategies
- Possibilistic learning in distributed and dynamic environments
- Adversarial learning and opponent modelling
- Collective learning and multi-agent adaptation
- Responsive learning
- Evolving interaction protocols and argumentation
- Direct and indirect interactions in coordination, competition, collaboration and coopetition
- Negotiated and emergent coalition formation, operation and disbanding
- Dynamic team composition and self-assembly
- Inverse design of dynamic interactions and decision strategies
- Smart Infrastructure (Smart Energy Grids, Smart Transport, Smart Internet): distributed management, optimal control and diagnosis (self-healing)
- Web/Grid/Cloud Services Management: service discovery, composition, SLA negotiation, QoS control and adaptation, service brokering
- E-commerce/E-business: coopetitive trading, dynamic contracting, business negotiation
- Virtual Enterprises and Organisations: dynamic formation, coordination and evolution of virtual organisations, integration of inter-organisational business processes, collaborative virtual environments and organisations
- Complex Adaptive Systems: modelling and optimisation of socio-economical systems
You can download our brochure in PDF by clicking here.