|
|||||
| Home | Projects | Publications | Repository |
|
|
| |||||
Below are full courses and guest lectures taught within regular University curricula. For summer-school and other intensive graduate-level courses, see also the Presentations section of the Repository.
Current courses
-
RL2024 - Reinforcement learning
2024-2025, 1st semester
Master CPS, Automation Department, UTCluj, Romania
A course on reinforcement learning (RL), geared mostly towards control engineers. Starts from discrete-space methods, continues with approximate RL, and puts significant emphasis on deep RL methods.
-
System identification 2024-2025
2024-2025, 1st semester
Control Engineering BSc, Faculty of Automation & Computer Engineering, UTCluj, Romania
Based on the book by Soderstrom and Stoica, the course introduces an array of methods for system identification, ranging from transient analysis, through prediction error methods, to recursive and closed-loop identification. It describes the material at an appropriate BSc level, and builds in a self-contained manner the required mathematical background, including statistics and linear regression. This course is taught within the English line of the BSc, as well as series B of the Romanian-language line. For the latter, there is a Romanian-language version of the course: Identificarea sistemelor.
-
CI2024 - Learning Control
2023-2024, 2nd semester
Master ICAF, Automation Department, UTCluj, Romania
A course on Learning Control that focuses mainly on reinforcement learning, starting from the basic methods in discrete spaces and ending with deep reinforcement learning.
Past courses (selection)
-
System identification 2023-2024
2023-2024, 1st semester
Control Engineering BSc, Faculty of Automation & Computer Engineering, UTCluj, Romania
Last year's System Identification course. Check this for a full set of lectures and labs (although the course evolves and some things will be different this year).
-
CI2023 - Learning Control
2022-2023, 2nd semester
Master ICAF, Automation Department, UTCluj, Romania
The previous edition of the course on Learning Control.
-
Optimization 2018
2017-2018, 2nd semester
Control Engineering BSc, Faculty of Automation & Computer Engineering, UTCluj, Romania
This course introduces optimization theory and numerical optimization techniques. It is taught within the 'A' profile of the English and Romanian BSc lines. For the latter line, there is a Romanian-language version of the course: Optimizari.
-
Guest lecture on Optimistic Optimization
The lecture presents two recent methods developed Remi Munos for the optimization of unknown functions under very general assumptions. This invited lecture is given yearly in the Optimization course taught by Zsofia Lendek in the Control Engineering BSc, at the Technical University of Cluj-Napoca, Romania. The title links to the latest version of the course material (slides), from May 2014.
-
Guest lecture on Learning Robot Control, within SC4240TU - Control Methods for Robotics
March 1st 2011. The lecture title links to the slides, and the course website is at http://www.dcsc.tudelft.nl/~sc4240tu.
-
SC4081 - Knowledge-Based Control Systems
2005-2011, 3rd quarter
System and Control Master, DCSC, TUDelft, the Netherlands
I supervised practical assignments and taught lectures in this course until 2010, while in 2011 I served in an advisory role. The reinforcement learning lectures I gave in 2010 can be downloaded from the Repository, where several demonstration movies used in the lectures are also available.
Other materials
A kickoff package for Master students starting their final project: student_package.zip. Since Matlab software is an important component of most projects, this kickoff package contains guidelines and a template for developing Matlab software, a miniassignment for quickly getting up to speed, and a document with tips on effectively searching for scientific papers. Another assignment focused more on simulation, control, and estimation aspects is available: controlassign.pdf.