Skip to content
Technische Universität Berlin Deutsch
Course catalog and module transfer system

Deep Learning 2Module description #41072 / Version 1

← Module page Deutsche Fassung
This module description is not valid in the current semester (Winter semester 2026/27). Current version (v2)
Module / Version#41072 / #1
ValidityWinter semester 2022/23 – Summer semester 2023
Available language(s)English
Credit points6
Person responsible for the moduleSchütt, Kristof
GradingGraded
Exam formatWritten exam
Language(s) of instructionEnglish

Affiliation

FacultyFakultät IV
InstituteInstitut für Softwaretechnik und Theoretische Informatik
Academic chair34352000 FG Maschinelles Lernen

Contact

OfficeFR 6-1
Contact personEberle, Oliver
E-mail addressoliver.eberle@tu-berlin.de
Websitehttp://wiki.ml.tu-berlin.de/wiki/

Learning outcomes

Understanding of advanced techniques of deep learning, including the application of mathematical concepts such as differential equations, group theory and fix-point iteration in neural networks. Understanding of advanced deep neural network architectures used in practice (e.g. GANs, Transformer, normalizing flows, equivariant neural networks). Ability to implement advanced neural networks using common deep learning frameworks.

Content

Advanced techniques such as self-attention, equivariant convolutions, deep equilibrium networks, generative adversarial networks, neural differential equations and normalizing flows. Application of machine learning techniques such as self-supervised / autoregressive learning, density estimation, anomaly detection and reinforcement learning in neural networks. Approaches to accelerate and deploy neural network, such as distillation, quantization, and factorization.

Module components

Compulsory area

The following courses are mandatory for this module:

CoursesTypeNumberFrequencyLanguageHours per weekCourse catalog
Deep Learning 2 - mainIntegrated course Summer semester[en]4 ISIS

Workload and credit points

Deep Learning 2 - main (Integrated course)

Workload descriptionMultiplierHoursTotal
Concepts & Theory106 h60 h
Exercises106 h60 h
Programming106 h60 h
180 h

The workload of the module adds up to 180 hours. The module therefore comprises 6 credit points.

Description of teaching and learning methods

The module will consist of weekly lectures accompanied with weekly homeworks.

Requirements for participation and examination

Desirable prerequisites for participation in the courses:

There are no formal prerequisites. However, the module builds on the module Deep Learning 1. Therefore basic knowledge of machine learning and neural networks is desirable. Prior knowledge multivariate calculus, probability theory and Python programming will be assumed.

Mandatory requirements for the module examination registration:

The examination requirements could not be retrieved.

Completion of the module

GradingGraded
Exam formatWritten exam
LanguageEnglish
Duration/Extent120 minutes

Duration of the module

The following number of semesters is estimated for taking and completing the module: 1 semester(s).

This module can be started in the following semesters: Summer semester.

Maximum number of participants

This module is not limited to a number of students.

Registration procedures

cf. course webpage

Recommended reading, lecture notes

Lecture notes

unavailable

Electronic lecture notes

unavailable

Recommended literature

No recommended literature given

Assigned programs

This module is not used in any program.

Miscellaneous

Not specified