Es ist eine neuere Version 2023W dieser LV im Curriculum Master's programme Artificial Intelligence 2024W vorhanden.
Workload
Education level
Study areas
Responsible person
Hours per week
Coordinating university
3 ECTS
B3 - Bachelor's programme 3. year
Mathematics
Dmitry Efrosinin
2 hpw
Johannes Kepler University Linz
Detailed information
Original study plan
Bachelor's programme Technical Mathematics 2012W
Objectives
A Markov chain is a mathematical model that is useful in the study of complex systems. The basic concepts of the Markov chain are the state of a system and the transition from one state to another. It is said that a system is in a certain state when random variables that fully describe the system take on the values assigned to that state. A transition of the system from one state to another occurs when the variables that describe the system change their values accordingly. The purpose of this course is to give an analytical structure to a Markov decision problem which at the same time describes the system sufficiently well and is still computationally usable.
Subject
Markov-chain with a discrete time
Controlled Markov-chain
Iterative solution for sequential decision processes
The policy-iteration for the solution of sequential decision processes
Applications of the policy-iteration algorithm
The policy-iteration algorithm for the processes with several ergodic classes
The sequential decision processes with discounting
Continuous-time Markov-chains
The controllable continuous-time Markov-chains
The continuous decision problems
The continuous decision problems with discounting
Conclusion
Criteria for evaluation
Written examination at the end of a semester
Methods
Slides and blackboard
Language
German
Study material
Script Howard R. Dynamic programming and Markov processes. Wiley Series, 1960. Puterman M. L. Markov decision process. Wiley series in Probability and Mathematical Statistics, 1994.
