Modulcode:	Inf-NNDL
Englische Bezeichnung:	Neural networks and deep learning
Modulverantwortliche(r):	Prof. Dr. Carsten Meyer
Turnus:	unregelmäßig (SS18 SS19 SS20 SS21 SS22 SS23 SS24 SS25)
Präsenzzeiten:	2V 2Ü
ECTS:	6
Workload:	30 h lectures, 30 h exercises, 120 h self studies
Dauer:	ein Semester
Modulkategorien:	BSc-Inf-WP (BSc Inf (21)) WI (BSc Inf (15)) MSc-Inf-WP (MSc Inf (21)) 2F-MEd-Inf-WP (MEd-Hdl Inf (21)) 2F-MA-Inf-WP (2F-MA Inf (21)) MSc-WInf-WP-Inf (MSc WInf (21)) PI (MSc Inf (15)) WI (MSc Inf (15)) WI (MSc WInf (15))
Lehrsprache:	Englisch
Voraussetzungen:

Kurzfassung:

Neural Networks and Deep Learning recently have gained strong interest (Deep Learning has been considered one of 10 breakthrough technologies by the MIT Technology Review 2013). The aim of the course is to provide a fundamental understanding of important concepts, algorithms, techniques and architectures of neural networks and deep learning.

Lernziele:

After completing the course, students should

• have a basic overview over neural network and deep learning concepts, algorithms and architectures, suitable applications, capabilities and limitations,
• be able to apply suitable neural network and deep learning techniques to new problems,
• analyze the outcome of neural network and deep learning experiments and explore potential methods to improve performance.

Lehrinhalte:

• Biological basis (neuron and networks)
• Artificial neuron models
• Artificial neural networks: Architectures and the learning problem
• Feedforward neural networks, multi-layer perceptron
• Learning in neural networks and the backpropagation algorithm
• Deep Learning: Motivation and concepts
• Convolutional neural networks
• (If time permits:) Recurrent neural networks: Long Short Term Memory (LSTM)
• (If time permits:) Unsupervised learning: Autoencoders
• (If time permits:) Generative models: Variational Autoencoder, Generative Adversarial Networks

Weitere Voraussetzungen:

• strong interest in neural networks and deep learning
• conceptual and analytical skills
• mathematical skills desired (linear algebra, analysis, calculus)
• programming skills desired (Python language)
• ability to work with software libraries (in Python) and Juypter notebook

Prüfungsleistung:

Written exam. Prerequisite for admission to the exam is to have passed 4 out of the first 5 lab exercise sheets.

Lehr- und Lernmethoden:

Beamer presentation and use of software tools.

Verwendbarkeit:

Literatur:

• Ian Goodfellow et al., "Deep Learning", MIT Press, 2016
• Michael Nielsen: "Neural Networks and Deep Learning", 2017

(More literature in the course)

Verweise:

Kommentar:

Students are asked to bring their own laptops to the laboratory classes. Laboratory assignments are encouraged to be solved in teams of maximally 4 students.