Michail G. Lagoudakis
Position: Associate Professor
Office: 145.A35
Phone: +30-28210-37244
Fax: +30-28210-37542
Email: lagoudakis intelligence.tuc.gr
Personal Page: http://www.lagoudakis.gr
Office: 145.A35
Phone: +30-28210-37244
Fax: +30-28210-37542
Email: lagoudakis intelligence.tuc.gr
Personal Page: http://www.lagoudakis.gr
Short CV
Michail G. Lagoudakis is an associate professor with the School of Electrical and Computer Engineering of the Technical University of Crete since September of 2005. Prior to this appointment, he held a postdoctoral fellow position at the Georgia Institute of Technology (September 2003 - June 2005), and a visiting researcher position at the Shannon Laboratory of AT&T Labs (May 2000 - August 2000). He received his doctoral degree in Computer Science with distinction from Duke University in May of 2003.Education
- 2003: Ph.D., Computer Science, Department of Computer Science,
Duke University,
Durham, NC, U.S.A.
Title:"Efficient Approximate Policy Iteration Methods for Sequential Decision Making in Reinforcement Learning". - 1998: M.Sc., Computer Science, Center For
Advanced Computer Studies,
University of Louisiana, Lafayette
(Formerly, University of Southwestern Louisiana - USL),
Lafayette, LA, U.S.A.
Title:"Mobile Robot Local Navigation with a Polar Neural Map". - 1995: B.Sc., Computer Science, Department of Computer Engineering and Informatics , University of Patras, Patras, Hellas (Greece)
Research Interests
- Machine Learning
Research in various aspects of machine learning with emphasis on reinforcement learning, whereby an agent learns how to act rationally in an unknown environment through trial and error.
[More information about Machine Learning]
- Robotics
Research in autonomous robotics with emphasis on probabilistic methods and algorithmic aspects. Our lab is the home of the RoboCup team Kouretes (www.kouretes.gr).
[More information about Robotics]
- Artificial Intelligence
Research in modern areas of Artificial Intelligence, including Game Playing, Intelligent Search, Probabilistic Reasoning, and Auction-Based Coordination.
[More information about Artificial Intelligence]
- RoboCup
Research in autonomous soccer-playing robots. Our lab is the home of the RoboCup team Kouretes (www.kouretes.gr).
[More information about RoboCup]
Teaching
- COMP 102: Structured Programming
Description
Complex applications of pointers in the C language. Pointers to pointers. Recursion. Introduction to Java and abstraction in object–oriented programming. The notion of a class and an object. Input/output, parameter passing in methods, access levels of member variables/methods/classes, overloading, inheritance, polymorphism, abstract classes. Abstract data types. Examples of abstract data types. Lists and their versions (single/double linked lists, circular lists). Queues and stacks. Divide and conquer strategies. Binary search trees. Hash–based structures. Simple sorting and search algorithms.
Courses Portal Link: Structured Programming
- COMP 402: Theory of Computation
Description
Sets, relations, alphabets, languages. Finite state automata, regular expressions, regular languages. Equivalence of finite automata and regular expressions. State minimization. Lexical analysis. Pushdown automata, context-free grammar, context-free languages. Equivalence of pushdown automata and context-free grammars. Syntactic parsing. Turing machines and extensions, unrestricted grammars, recursive languages. Non-determinism, non-deterministic Turing machines, recursive enumerable languages. The language hierarchy. Decidability, computability, non-computability. Church-Turing thesis. Universal Turing machines, reductions. Rice's theorem. Computational complexity and complexity classes. Cook's theorem. Application to compiler construction and laboratory instruction of the tools flex, bison, JavaCC.
Courses Portal Link: Theory of Computation
- COMP 417: Artificial Intelligence
Description
Foundation and history of Artificial Intelligence. Intelligent agents and environments. Systematic search methods: uninformed, informed, heuristic. Local search methods. Constraint satisfaction problems and algorithms. Basic game theory and adversarial search. Propositional logic, first-order logic, reasoning, inference algorithms. Knowledge representation and knowledge bases. Reasoning systems, theorem provers, logic programming. Planning problems and algorithms. Planning in the real world and multi-agent planning.
Courses Portal Link: Artificial Intelligence
- COMP 513: Autonomous Agents
Description
Agents and environments, uncertainty and probability, probabilistic reasoning. Bayesian networks, exact and approximate inference in Bayesian networks, enumeration and sampling algorithms. Temporal probabilistic reasoning (filtering, prediction, smoothing, most likely sequence), dynamic Bayesian networks. Mobile robot navigation: motion control, path planning, localization, mapping, simultaneous localization and mapping (SLAM). Decision making under uncertainty, Markov decision processes, optimal policies, value iteration, policy iteration, partial observability. Reinforcement learning, prediction and control, basic and advanced reinforcement learning algorithms. Approximation methods for multi-dimensional and continuous spaces. Competitive agents, planning and learning in Markov games. Auction-based multi-agent coordination. Applications to autonomous robotic agents and laboratory instruction of robot programming tools.
Courses Portal Link: Autonomous Agents
- COMP 604: Machine Learning
Description
Basic concepts of machine learning and statistics. Supervised learning: least mean squares (LMS), logistic regression, perceptron, Gaussian discriminant analysis, naive Bayes, support vector machines, model selection and feature selection, ensemble methods (bagging, boosting). Learning theory: bias/variance tradeoff, union and Chernoff/Hoeffding bounds, VC dimension. Unsupervised learning: clustering, k-means, EM, mixture of Gaussians, factor analysis, principal components analysis (PCA), independent components analysis (ICA). Reinforcement learning: Markov decision processes (MDPs), Bellman equations, value iteration, policy iteration, value function and policy approximation, least-squares methods, reinforcement learning algorithms, partially observable MDPs (POMDPs), algorithms for POMDPs.
Courses Portal Link: Machine Learning
- COMP 614: Probabilistic Robotics
Description
Uncertainty and probabilistic reasoning. Robotics perception and action. Recursive state estimation: state space, belief space, prediction and correction, Bayes filter. Estimation filters: linear Kalman filter, extended Kalman filter, unscented Kalman filter, histogram filter, particle filter. Probabilistic motion models: velocity model, odometry model, sampling and density. Probabilistic observation models: beam model, scan model, feature model, sampling and density. Robot localization: Markov, Gaussian, Grid, Monte-Carlo. Robotic mapping: occupancy grid maps, feature maps, simultaneous localization and mapping (SLAM). Decision making under uncertainty, Markov decision processes, optimal policies, value iteration, policy iteration, partial observability. Reinforcement learning, prediction and control, basic and advanced reinforcement learning algorithms. Multi-robot coordination and learning.
Courses Portal Link: Probabilistic Robotics
Undergraduate Courses
Postgraduate Courses
Supervision
Publications
Funding
- NOPTILUS: autoNomous, self-Learning, OPTImal, and compLete Underwater Systems
Design and implementation of a fully autonomous team of underwater vehicles for exploratory missions.
[More information about this project] - Kouretes: Technical University of Crete RoboCup Team Kouretes
Development of a fully-autonomous soccer team of humanoid robots Nao
[More information about this project] - CALR: Coordinated Action and Learning in RoboCup
This research studies fundamental problems of robotics, machine learning, and artificial intelligence in the context of the international RoboCup competition for robotic soccer. In particular, it focuses on coordinated action and learning aiming at enabling better coordination among the team members.
[More information about this project] - AuctionCoord: Evaluation of Auction-Based Algorithms for Coordination of Robot Teams
Experimental Evaluation of Auction-Based Distributed Algorithms for Cooperative Coordination of Robot Teams
[More information about this project] - RLvSL: Reinforcement Learning via Supervised Learning
This project investigates the potential of using supervised learning technology for advancing reinforcement learning. It is possible to incorporate supervised learning algorithms within the inner loops of several reinforcement learning algorithms and therefore reduce one problem to the other. This synergy opens the door to a variety of promising combinations. This research will establish the criteria under which this reduction is possible, will investigate viable combinations, will propose novel algorithms, will assess their potential, and will apply them to real problems of practical interest to demonstrate their effectiveness.
[More information about this project]