Lecture: Probabilistic Graphical Models (Gastmodul Bioinformatik A) - Details
You are not logged into Stud.IP.

General information

Course name Lecture: Probabilistic Graphical Models (Gastmodul Bioinformatik A)
Semester SS 2013
Current number of participants 4
Home institute Bioinformatik
participating institutes Institut für Informatik
Courses type Lecture in category Offizielle Lehrveranstaltungen
First date Monday, 24.06.2013 10:00 - 14:00
Pre-requisites Molekulare Phylogenie
Performance record - Scheinerwerb durch Lösung der Aufgaben und aktive Teilnahme an den Übungen
- mündliche Prüfung
Studiengänge (für) Master Bioinformatik

Rooms and times

No room preference
Monday: 10:00 - 14:00, weekly(3x)


In the last twenty years, Probabilistic Graphical Models (PGMs) have emerged as a most effective tool for solving artificial intelligence problem in general and in particular for bioinformatics applications. This course introduces what PGMs are and reviews the different types of questions that we can answer through its use. We cover algorithms that allow us to provide an exact answer to such questions, and establish when this answer can be provided in polynomial time. For those cases when no polynomial time exact algorithm exists, we review approximate inference techniques. In particular, we concentrate on structured variational approaches and variational expectation maximization. Along the course, exercise sessions will be used to show how the theoretical results in the lectures can be applied to particular bioinformatics problems, with an emphasis in molecular phylogenetics.