Prof. Dr. Moritz Diehl
Modeling and System Identification (MSI) is concerned with the search for mathematical models for real-life systems. The course is based on statistics, optimization and simulation methods for differential equations. The exercises will be based on pen-and-paper exercises and computer exercises with MATLAB.
Lectures take place on Wednesdays 08:30h to 9:55h and Fridays 10:05 to 11:50, in HS 00 036 (Schick - Saal) in building 101.
Recordings of some of the lectures are available on the webpage of the video center.
Course material is the following:
- MSI script (updated October 23, 2019) by Prof. Diehl,
- Script by Prof. Johan Schoukens, VUB, Brussels, Belgium,
- Textbook, Ljung, L. (1999). System Identification: Theory for the User. Prentice
Hall. Available in the campus library.
Tentative course schedule (may change, please check regularly):
|October 23||Complete lecture|
|October 25||Linear Algebra Tutorial|
|October 30||Complete lecture|
|November 6||Statistics Tutorial||Complete tutorial|
|November 8||Complete lecture|
|November 13||Complete lecture|
|November 15||Complete lecture|
|November 20||Complete lecture|
|November 22||Complete lecture|
|December 4||Complete lecture|
|December 6||Microexam 1 & Solution||Complete lecture|
|December 13||No new recordings for this winter term.
Please watch the old recordings from 2017/18.
|December 20||no lecture|
Machine Learning in a Nutshell
|January 22||Complete lecture|
|January 24||Microexam 2||Complete lecture|
|January 31||Complete lecture|
|February 5||Complete lecture|
|February 7||no recordings|
|February 12||Microexam 3||no recordings|
|February 14||Summary Lecture||no recordings|
Exercise sessions are organized on (starting on October 24, 2019):
- Thursday 16:00 to 18:00
- Friday 12:00 to 14:00
- Tuesday 12:00 to 14:00
in building 082, room 029.
Please hand in solutions to computer exercises through Matlab Grader individually (you should have received an invitation email, email us). Solutions to non-computer exercises can be handed in on paper by groups of maximum 3 persons during the Wednesday lecture or before that in building 102, 1st floor, 'Anbau' (here). The corrected exercises will be handed out during the exercise sessions.
- Exercise 0 (updated)
- Exercise 1 dataset
- Exercise 2 (updated)
- Exercise 3 dataset
- Exercise 4 (updated)(updated) dataset
- Exercise 5 dataset
- Exercise 6 dataset
- Exercise 7 dataset
- Exercise 8 (updated) dataset_task1 dataset_task2
- Exercise 9 (updated)(updated)
- Exercise 10 dataset
In order to pass the exercises accompanying the course (`Studienleistung`), one has to obtain at least 20 exercise points in each of the three blocks:
- Block: Exercises 0 - 3 + Microexam 1,
- Block: Exercises 4 - 7 + Microexam 2, and
- Block: Exercises 8 - 11 + Microexam 3.
After each Microexam we will provide an anonymous list of the number of exercise points as well as the result of the microexam of each student here (1. Block). If you are interested in your current number of exercise points, send us an email at any time or ask us at the exercise sessions.
If you have any questions regarding the exercises, email us.
- Tobias Schöls
- Jia-Jie Zhu
- Naya Baslan
- Jakob Harzer
- Bryan Ramos
If you have questions please approach us during the exercise sessions. In urgent cases you may also send an email to firstname.lastname@example.org
The final exam will take place on March 20, 2020 at 14.00h in lecture halls 026 + 036 in building 101.
The material for the tutorials:
We recommend students to install MATLAB on their laptop and bring it to the exercise sessions (and the tutorials in the beginning of the semester). The university provides licences.
There is an online (in browser) version of MATLAB. This service is provided by MathWorks and can be accessed with a MathWorks account. We won't be able to provide support for the online version and the exercises may exceed its capabilities.Ultimately MATLAB is installed on some come computers in the computer pool. We have no influence on this installation please refer to the pool managers for details.
EXTRA EXAMPLES AND RIDDLES
- THE MOVING BLACKBOARD RIDDLE: The plot and the corresponding MATLAB datafile shows the recorded values of control input values (-1,0,1) for a (virtual) electrically actuated blackboard for a timespan of 60 seconds on the top plot. The lower plot shows measurements of the height (i.e., the output) of the blackboard for the first 40 seconds (in meters). QUESTION: Model the dynamics of the system, identify the relevant parameters, and predict the height of the blackboard at time t=60 s.
- MACHINE LEARNING OPTIONAL EXERCISE SHEET : (Jupyter Notebook) msi-ml.ipynb . This is an optional exercise sheet that will provide an introduction to machine learning using Python.