Algorithmic Decision Theory: Combining Decision Analysis and Artificial Intelligence: Theoretical Framework and Practical Applications
There is increasing awareness among the AI community about the importance of Decision Analysis concepts, methods and models for the development of intelligent, autonomous systems. At the same time Decision Analysts are increasingly aware that findings in Artificial Intelligence are becoming extremely important in order to handle scientific and operational challenges in supporting decision makers and/or in developing decision support devices in many areas of application. A large community of researchers and practitioners is today attracted by the synergy the two fields provide both in scientific investigation and real life applications from homeland security to the semantic web, from e-government to recommender systems.
The tutorial aims to introduce a general framework within which to show the large variety of theoretical achievements in Decision Analysis (measurement theory, preference modeling, optimisation, social choice theory) and how these are relevant for AI research. It also aims to show the importance of joint research findings both in knowledge advances (preference learning, uncertainty and robustness in decision making, algorithmic aspects of decision support) and in real life decision support applications. The tutorial is a result of ongoing discussions within the COST Action IC0602 on "Algorithmic Decision Theory", funded by the European Union (see www.algodec.org).
Alexis Tsoukiàs (1959) is a CNRS Research Director at LAMSADE a research center specialised in Decision Support and Computer Science. He specializes in Decision Analysis, Operational Research, Preference Modeling and their relations to Artificial Intelligence. He published two books and more than 70 papers on these subjects. He teaches Decision Theory, Logic, and Evaluation Models. He is a Past President of EURO (The European Association or Operational Research). He organized in 2001, the EURO Summer Institute in Decision Analysis and Artificial Intelligence. Since 2004 he is co-principal investigator of the LAM SADE -DIMACS project on Algorithmic Decision Theory and since 2007 he is the Chair of the COST Action IC0602 on Algorithmic Decision Theory (funded within the FP7 of the EU). He has been invited worldwide by Universities and Research centres for workshops, seminars and tutorials.