Michail G. Lagoudakis was born in 1972 in Irakleio, Crete in Greece.
In 1990, he enrolled in the Department of Computer Engineering and Informatics at the University of Patras in Greece. He completed the five years program constantly ranking in the top 10 of his class (a total of 120) and he was awarded a scholarship by the National Scholarship Foundation of the Greek Ministry of Education. His diploma thesis, entitled Implementation of a Knowledge-Based Scheduler for Job-Shop Production Environments (coauthored with his colleague Nikolaos Parlavantzas), was supervised by Prof. Paul Spirakis. This work was part of the larger European ESPRIT CIM Project at the University of Patras. He graduated in July, 1995 and received the Diploma of Computer and Informatics Engineer with a grade "Excellent" (8.67/10.00) ranking 10th among his class. During his military service he served for five months as the Database Administrator for the Artillery Training Center in Greece.
In August 1996 he enrolled in the graduate program of the Center for Advanced Computer Studies at the University of Louisiana, Lafayette. At the same time he was awarded a graduate scholarship from the Lilian-Boudouri Foundation in Greece. He graduated with the M.Sc. degree in Computer Science in May 1998, and during the course of study he was honored twice by the USL Honors Program. His M.Sc. thesis, entitled Mobile Robot Local Navigation with a Polar Neural Map, was completed under the supervision of Prof. Anthony S. Maida and was used for sonar-based navigation of a Nomad 200 mobile robot at the Robotics and Automation Laboratory directed by Prof. Kimon Valavanis.
In August 1998, he enrolled as a graduate student in the Department of Computer Science at Duke University. He received a graduate student scholaship from Duke University and the Outstanding Teachning Assistant Award twice. His Ph.D. research was supervised initially by Prof. Michael Littman and later by Prof. Ronald Parr. His dissertation, entitled Efficient Approximate Policy Iteration Methods for Sequential Decision Making in Reinforcement Learning, proposed two new efficient algorithms for the problem of sequential decision making with extensions to multiagent and/or competitive domains. His dissertation received the departmental Outstanding Disseration Award for the academic year 2002-2003.
His research interests fall in Numeric Artificial Intelligence. More particularly he is interested in machine learning, especially reinforcement learning, Markov decision processes and optimal control. He also maintains a strong interest in artificial and biological neural systems with focus on sensory-motor control.
He is a member of the American Association for Artificial Intelligence (AAAI), the Institute of Electrical and Electronics Engineers (IEEE), and the Association for Computing Machinery (ACM).