Foundations of Agents introduces the student to the formal foundation of agents. An agent is a computational being, such as a software program, robot or human. Agents operate in some environment, which they can observe and in which they can realize objectives through the execution of actions. Examples of such environments are computer game environments, the internet and the physical world in case of robots and humans. This course addresses the problem of how an agent can act optimally in order to realize its objectives. It is investigated how the agent's environment, its objectives, actions and observations can be formalized to solve this problem. Initially Markov decision processes will be used to model the agent and its environment. Other models that will be investigated include logic-based models, such as epistemic logic, doxastic logic, dynamic logic, temporal logic, and BDI logics. The extension of these models towards multi-agent systems will be emphasized as often agents do no not operate alone in an environment. After completing this course the student will be familiar with the formal models of describing agents and how these models can be used to determine and agent's behaviour. The student will be able to judge which model is applicable for specific problem domains.
Knowledge of Propositional and Predicate Logic, Calculus