Konvexe Optimierung (Convex Optimization)

Lecturer
Prof. Dr. Frank Vallentin

Coordination of the Exercise Sessions
Dr. Anna Gundert, Dr. Frederik von Heymann

Sprache/Language
Übungsaufgaben und die Abschlussklausur können wahlweise in Englisch oder in Deutsch bearbeitet werden.
The course will be organized in English. Exercises and the final exam can be submitted either in German or in English.

Contents
In modern „Convex Optimization“ the theory of semidefinite optimization plays a central role. Semidefinite optimization is a generalization of linear optimization, where one wants to optimize linear functions over positive semidefinite matrices restricted by linear constraints. A wide class of convex optimization problems can be modeled using semidefinite optimization. On the one hand, there are algorithms to solve semidefinite optimization problems, which are efficient in theory and practice. On the other hand, semidefinite optimization is a tool of particular usefulness and elegance.

The aim of this course is to provide an introduction to the theory of semidefinite optimization, to algorithmic techniques, and to mathematical applications in combinatorics, geometry and algebra.

Lectures
Großer Hörsaal der Mathematik (Raum 203)
Wednesday 12.00-13.30
Friday 8.00-9.30

Note: The lecture on Friday, Dec 21, will not take place. Please study this week’s material (Chapter 7, § 5) on your own. It will be relevant for the oral exam.

Exercise Sessions
Please see the Exercises-page for more information.

Exam date
Wednesday 30.01.2019, 12.00-13:30 (Raum 203 MI)

Results of the Exam
Here. Post-exam review on Monday, Feb. 4, 14:30-15:30 in our offices (Weyertal 80).

Example Questions for Oral Exams
Here

Appointments for the Oral Exams
You can find the file with the appointments here (If you cannot remember the password, send us a mail or come to our office hours).
Signatures for the registration form can also be obtained from Annette Koenen (Weyertal 80, 1st floor on the right).

Practice Exam
To get an idea of how an exam could look like, here is an exam from 2016. Of course the material is not identical so this should not be taken as a definitive selection of topics for exam questions.