Machine Learning & Deep Learning
Niveau
Master's course
Learning outcomes of the courses/module
The following skills are developed in the course:
- Students are familiar with tools (e.g. libraries, cloud platforms or software tools), with which machine learning can be supported.
- Students can compare the tools developed with regard to their suitability for specific problems.
- Students can design end-to-end machine learning projects.
- Students can carry out end-to-end machine learning projects independently
- Students are familiar with tools (e.g. libraries, cloud platforms or software tools), with which machine learning can be supported.
- Students can compare the tools developed with regard to their suitability for specific problems.
- Students can design end-to-end machine learning projects.
- Students can carry out end-to-end machine learning projects independently
Prerequisites for the course
1st semester: Students have previous knowledge of mathematics/statistics up to 8 ECTS and therefore know simple statistical measures as well as basic statistical test procedures (e.g. t-test). / 2nd semester: No prerequisites / 2nd semester: Module examination MLAL.A1 (Algorithmic 1)
Course content
The following content is discussed in the course:
- Classical neural networks as a supplement to classical algorithms of data science (e.g. Random Forests, SCM, etc.)
- Fallen, artificial neural networks (CNN)
- Recursive, artificial neural networks (RNN, LSTM)
- Continuing, artificial neural networks (GAN, FARM, BERT, CGAN, etc.)
The network types discussed are subject to constant change. For this reason, only a few network types are men-tioned here as examples. Current network types are also discussed and applied in the course.
- Classical neural networks as a supplement to classical algorithms of data science (e.g. Random Forests, SCM, etc.)
- Fallen, artificial neural networks (CNN)
- Recursive, artificial neural networks (RNN, LSTM)
- Continuing, artificial neural networks (GAN, FARM, BERT, CGAN, etc.)
The network types discussed are subject to constant change. For this reason, only a few network types are men-tioned here as examples. Current network types are also discussed and applied in the course.
Recommended specialist literature
PRIMARY LITERATURE:
- Géron, A. (2017): Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and Techniques for Building Intelligent Systems (Ed. 1), O´Reilly, Farnham (ISBN: 978-1491962299)
- Géron, A. (2017): Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and Techniques for Building Intelligent Systems (Ed. 1), O´Reilly, Farnham (ISBN: 978-1491962299)
Assessment methods and criteria
Project documentation and presentation
Language
English
Number of ECTS credits awarded
10
Share of e-learning in %
25
Semester hours per week
4.0
Planned teaching and learning method
The following methods are used:
- Processing of exercises
- Interactive workshop
- Processing of exercises
- Interactive workshop
Semester/trimester in which the course/module is offered
2
Name of lecturer
Dr. Michael Hecht, Prof. (FH) Dr. Michael Kohlegger
Academic year
Key figure of the course/module
MLAL.3
Type of course/module
integrated lecture
Type of course
Compulsory
Internship(s)
none