ELL 705, II semester 2016-17
This is the the webpage for ELL 705, II semester 2016-17
Stochastic Filtering and Identification
Instructor
Shaunak Sen
E-mail: shaunak.sen@ee.iitd.ac.in
Lectures: MTuF 12:00-12:50pm
Room: IIA-201
Announcements
28.04.2017: MAJOR TEST scheduled from 3:30PM-5:30PM on Sat, 05.05.2015 in LH621. Please check Exam Schedule for up-to-date information.
17.04.2017: MINOR TEST 2 solutions posted.
22.03.2017: MINOR TEST 2 scheduled from 11AM-12PM on Sat, 25.03.2017 in LH623. Please check Exam Schedule for up-to-date information.
06.03.2017: HW2 and Project information posted.
20.02.2017: MINOR TEST 1 solutions posted.
30.01.2017: MINOR TEST 1 scheduled from 2-3PM on Fri, 03.02.2017 in LH623. Please check Exam Schedule for up-to-date information.
24.01.2017: HW1 posted.
14.01.2017: Webpage online.
Lectures
| S. No. | Date | Topic | Advised Reading | Homework |
| 1 | Week 1
(Jan 3) | Introduction | Ljung 1.4 | |
| 2 | Week 2
(Jan 10, 13) | Example (Least-Squares)
Mean and Variance of Estimate | Lecture
Ljung 1.3 | |
| 3 | Week 3
(Jan 16, 17, 20) | Example (Kalman Filter)
Models
Linear Discrete-Time Models
(ARX, ARMA, ARMAX, Box-Jenkins)
Predictors
Others – State-Space, Linear Time-Varying, Nonlinear
Methods
Least-Squares |
Ljung 4.1-4.2
Ljung 3.1-3.2
Ljung 4.3, 5.1-5.3
Lecture | |
| 4 | Week 4
(Jan 23, 24, 27) | Least-Squares
Discussion | Ljung Appendix II, 7.4
Lecture, Lecture | HW1 |
| 5 | Week 5
(Jan 30, Feb 3) | Least-Squares
MINOR TEST 1 | Lecture | |
| 6 | Week 6
(Feb 6, 7, 10) | Pseudo-Linear Regression
Instrument-Variable Methods
Maximum Likelihood | Lecture, Code
Ljung 7.6 Lecture, Code
Ljung 7.4, Lecture | |
| 7 | Week 7
(Feb 13, 14) | Cramer-Rao Bound
Recursive Least-Squares | Ljung Appendix 7A, Lecture
Ljung 11.1-11.5, Lecture | |
| 8 | Week 8
(Feb 20, 21, 22) |
Time-Varying Parameters
Convergence | Lecture
Lecture
Lecture, Ljung 8.1-8.4, 9.1-9.3, Code | |
| 9 | Week 9
(Mid-Semester Break) | | | |
| 10 | Week 10
(Mar 6, 7, 10) | Extended Kalman Filter
Define Parameter as State
Linearization | Anderson & Moore 1, 8.1-8.2
Lecture, Lecture, Lecture | HW2
Project |
| 11 | Week 11
(Mar 14, 17) | Conditional Probability
Moment Generating Functions | Anderson & Moore 2.3
Lecture, Lecture | |
| 12 | Week 12
(Mar 20, 21, 25 - Minor Test 2) | Minimum Variance Estimate
HW2 discussion | Anderson & Moore 2.3
Lecture, Lecture | |
| 13 | Week 13
(Mar 27, 28, 31) | Discrete-time Kalman Filtering
Problem and Solution | Anderson & Moore 3.1
Lecture, Lecture, Lecture | |
| 14 | Week 14
(Apr 1, 3, 7) | Proof | Anderson & Moore 3.1 | |
| 15 | Week 15
(Apr 11) | Properties - BLUE, Time-Invariance, Stability | Anderson & Moore 3.2, 4
Lecture | |
| 16 | Week 16
(Apr 17, 18, 21) | Linear Minimum Variance Estimator
Innovation Sequence | Anderson & Moore 5
Lecture, Lecture | |
| 17 | Week 17
(Apr 25, 27, 28) | Projects
Particle Filter Overview | |
|
Reference Textbooks
L. Ljung, System Identification: Theory for the User, Prentice Hall, 2nd edition, 1999.
T. Soderstrom and P. Stoica, System Identification, Prentice-Hall, 1989.
B. D. O. Anderson and J. B. Moore, Optimal Filtering (Dover Books on Electrical Engineering).
|