Parameter estimation - Paraméterbecslés

Parameter estimation - Paraméterbecslés

(VEMIVIM133P, MI MSc)

1st semester of 2017/2018

Lecture-tutorial-laboratory: Wednesday 11:15-14:00, I/207B

Lecturer: Prof. Katalin Hangos ( Ezt a címet a spamrobotok ellen védjük. Engedélyezze a Javascript használatát, hogy megtekinthesse. )
Tutor: Anna Ibolya Pózna ( Ezt a címet a spamrobotok ellen védjük. Engedélyezze a Javascript használatát, hogy megtekinthesse. )

Course evaluation

The evaluation is based upon a mid-semester closed-book exam and on a parameter estimation project work to be implemented in MATLAB.
The pre-requisite of the course signature is
-- to be present at least 75% of the lectures-tutorials-laboratories, 
-- to submit the project results and documentation to the given deadline, and
-- to achieve at least 50% on the closed-book exam.

Course contents

Week Date Type Material Slides/supplements
1 6 Sept
Registration week
2 13 Sept Lec/Tut Recalling the basic knowledge on scalar and vector valued random variables, their properties and independence. The basics of mathematical statistics: sample, statistics-estimation. The estimation of mean value, covariance and correlation.

icon PB_basic_notions

icon PBtut_basic_notions

3 20 Sept Lab MATLAB basics. Simple functions for estimating mean value, covariance and correlation.

icon PB_Lab_1

icon autocov (131 Bytes)

icon PB_Lab_1_Matlab (3.2 kB)

4 27 Sep Lec/Tut The properties of the estimates. Parameter estimation of static models. Linear regression and its properties.

icon PB_linear_regression

icon PBtut_linear_regression

5 4 Oct Lab Linear regression and its properties.

icon datasets (7.14 kB)

icon PB_Lab_2 (167.18 kB 2017-10-04 16:24:11)

icon leastsquares (631 Bytes)

6 11 Oct Lec/Tut Discrete time stochastic processes, the input-output models of discrete time stochastic linear time-invariant systems. The principle of parameter estimation in dynamic case.

icon PBLinDynMod

icon PBtut_Linear_Dynamic_Models

7 18 Oct Lec/Tut Methods that are based on minimization of the prediction error. Methods based on least squares (LS) minimization. The properties of the LS estimation method. icon PB_DynLS
8 25 Oct Lab Methods based on least squares (LS) minimization.
9 8 Nov Lec/Tut The maximum likelihood (ML) estimation. ML estimation of the parameters of predictive models. The covariance matrix of the estimate. Parameter estimation for nonlinear models. The instrumental variable (IV) method.
10 15 Nov Lab Parameter estimation for nonlinear models. The instrumental variable (IV) method.
Project work specification

12 29 Nov Lec/Tut The practical implementation of parameter estimation. Diagnosis based on parameter estimation.
Consultation

13 6 Dec Lec/Tut
Lab
CLOSED BOOK EXAM
Consultation for the project work

14 13 Dec Lab Submission DEADLINE of the PROJECT WORK

Keresés a honlapon