Zur Merkliste hinzufügen
Zurück

 

67923-01 - Vorlesung: Continuous Optimization 6 KP

Semester Frühjahrsemester 2023
Angebotsmuster unregelmässig
Dozierende Aurelien Lucchi (aurelien.lucchi@unibas.ch, BeurteilerIn)
Inhalt This course provides an in-depth theoretical treatment of classical and modern optimization methods. Specifically, we will discuss the following concepts:

Basic convex analysis
Subgradient methods
Gradient Descent
Convergence rate for gradient-based methods
Optimality and lower bounds
Stochastic Optimization methods
Non-convex Optimization
Lernziele - Equip students with a fundamental understanding of why optimization algorithms work, and what their limits are
- Ability to select optimization algorithms for practical applications that students might encounter in their future career
- Ability to understand and derive mathematical proofs for optimization algorithms
Literatur The first part of the class will cover some of the chapters discussed in the following books:
Numerical Optimization, by Jorge Nocedal and Stephen J. Wright
Convex Optimization: Algorithms and Complexity, by Sebastian Bubeck
Convex Optimization, by Stephen Boyd and Lieven Vandenberghe

The second part of the class will mostly be based on research papers.
Bemerkungen Exceptionally, the first lecture will be given on Monday, February 20 from 4.15pm to 6pm. There won't be any lecture on Thursday, February 23rd.

Starting week 2: Exercise sessions will start the second week of the semester and will be scheduled every Monday (4.15pm to 6pm). The lectures will be scheduled on Thursdays from 4.15pm to 6pm.

 

Teilnahmebedingungen A solid background in analysis and linear algebra; some background in theoretical computer science (computational complexity, analysis of algorithms); the ability to understand and write mathematical proofs.
Unterrichtssprache Englisch
Einsatz digitaler Medien kein spezifischer Einsatz

 

Intervall wöchentlich
Datum 23.02.2023 – 01.06.2023
Zeit Donnerstag, 16.15-18.00 Spiegelgasse 1, Seminarraum 00.003
Datum Zeit Raum
Donnerstag 23.02.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 02.03.2023 16.15-18.00 Uhr Fasnachstferien
Donnerstag 09.03.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 16.03.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 23.03.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 30.03.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 06.04.2023 16.15-18.00 Uhr Ostern
Donnerstag 13.04.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 20.04.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 27.04.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 04.05.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 11.05.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 18.05.2023 16.15-18.00 Uhr Auffahrt
Donnerstag 25.05.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Donnerstag 01.06.2023 16.15-18.00 Uhr Spiegelgasse 1, Seminarraum 00.003
Module Modul: Applications of Distributed Systems (Masterstudium: Computer Science)
Modul: Applications of Machine Intelligence (Masterstudium: Computer Science)
Modul: Concepts of Machine Intelligence (Masterstudium: Computer Science)
Modul: Methods of Machine Intelligence (Masterstudium: Computer Science)
Leistungsüberprüfung Lehrveranst.-begleitend
Hinweise zur Leistungsüberprüfung Continuous assessment

Note the following split:
30% homework
30% project (writeup and presentation)
40% written exam

A 50% score on HW sets is required to participate in the final exam.

Expected date: Monday, 10 July 2023, 10-12 a.m., Spiegelgasse 1, room 00.003.
An-/Abmeldung zur Leistungsüberprüfung An-/Abmelden: Belegen resp. Stornieren der Belegung via MOnA
Wiederholungsprüfung keine Wiederholungsprüfung
Skala 1-6 0,5
Wiederholtes Belegen beliebig wiederholbar
Zuständige Fakultät Philosophisch-Naturwissenschaftliche Fakultät, studiendekanat-philnat@unibas.ch
Anbietende Organisationseinheit Fachbereich Informatik

Zurück