Algorithmik des maschinellen Lernens für Graphen

Modulnummer: Q10-37
Englischer Titel: Algorithmic Machine Learning Methods for Graphs
Leistungspunkte: 10
Lehrperson: Meyerhenke

Empfohlene Vorkenntnisse

Good knowledge of fundamental graph algorithms as well as linear algebra operations
(as taught in Bachelor modules similar to "Algorithms and Data Structures" and
"Linear Algebra and its Connections to Computer Science") is strongly recommended.

Zwingende Voraussetzungen

none

Inhalt

A graph is a versatile data structure that can represent complex data as relationships
between objects. Real-world networks modeled by graphs occur in many applications.
Their analysis is key to understanding the structure and dynamics of the modeled processes.
This course focuses on the algorithmic foundations of the analysis of massive graphs with
machine learning and data mining methods.

A key learning objective is to master the full algorithm engineering cycle in the
context of the course. This means to model a real-world problem as algorithmic task,
to design algorithmic solution methods for the task, to be able to analyze and compare
these solution methods on a theoretical and empirical level, to implement (selected)
algorithms, and to design and evaluate systematic experiments.

Topics include representation learning, graph neural networks, graph clustering,
link prediction, network motifs, and others.

Erforderliche Arbeitsleistungen für LP-Vergabe und Prüfungszulassung

- schriftlich eingereichte und/oder mündlich vorgetragene Lösungen zu Aufgaben

Lehrveranstaltungen

Vorlesung: 4 SWS 6 LP
Übung: 2 SWS 3 LP
MAP: 1 LP

Zugeordneter Vertiefungsschwerpunkt

Algorithmen und Modelle: ja
Modellbasierte Systementwicklung: nein
Daten- und Wissensmanagement: ja
Ohne Vertiefungsschwerpunkt: nein

Sprache im Modul

Deutsch: nein
Englisch: ja

Angeboten für Studiengänge

M. Sc.: ja
M. Ed.: ja
Wirtschaftsmaster: ja

Angeboten im

Wintersemester: nein
Sommersemester: ja

Turnus

Unregelmäßig