EE557 ESTIMATION THEORY
Course Code: | 5670557 |
METU Credit (Theoretical-Laboratory hours/week): | 3 (3.00 - 0.00) |
ECTS Credit: | 8.0 |
Department: | Electrical and Electronics Engineering |
Language of Instruction: | English |
Level of Study: | Graduate |
Course Coordinator: | Prof.Dr. UMUT ORGUNER |
Offered Semester: | Fall Semesters. |
Course Objectives
At the end of this course students will learn
- Basics of stochastic discrete time systems given in state space form
- Markov Decision Processes
- Kalman Filtering
- Solution of LQG problem
Course Content
Gauss-Markov process and stochastic differential equations. Bayesian estimation theory. Maximum likelihood, linear minimum variance and least-square estimations. Properties of estimators; error analysis. State estimation for linear systems, Kalman-Bucy and Wiener filters. Smoothing and prediction. Nonlinear estimation. Filter implementation. Applications to communication, control, system identification and biomedical engineering.
Course Learning Outcomes
- Basics of linear stochastic systems that evolve in a Markovian fashion
- Details of finite state Markov Chains and controlled Markov chains
- Dynamic programming to be used in the Markov Decision Processes
- Markov Decision Processes restricted to the finite state Markov chain case
- Use of dynamic programming in Markov Decision Processes
- Meaning of partial information
- Kalman filtering as a state estimation tool for linear Gaussian systems
- Solution of the LQG problem as a stochastic partial observation optimal control problem