Module Data Science, Computer Science (Master) (ER 8)
Module summary
Data Science
INFM120ML
Prof. Dr. Reimar Hofmann
/
All semesters
none
none
At the end of this module, students will have understood the theoretical and mathematical foundations of machine learning and data analysis. They will be able to assess the suitability of different methods for specific situations, interpret phenomena they observe conclusively and, if necessary, derive ideas for improving the selected approaches.
The skills taught in the module are advantageous for participation in the Artificial Intelligence module.
Written/verbal Exam 120/20 Min. (graded)
Course Data Science
INFM121ML.a
Lecture
Prof. Dr. Reimar Hofmann
German
2/2
60 hours in total, including 30 hours of contact study.
Module exam
- Scale types (nominal, ordinal, interval, ratio), conversion of scale types (one-hot-coding etc.), standardization
- Explorative data analysis, differentiation between direct and indirect dependencies, data visualization
- Statistical principles of machine learning, maximum likelihood approach, bias and variance (overfitting) as sources of learning error
- Cost functions for numerical regression and classification
- Criteria foe data quality, dealing with quality deficiencies (e.g. missing values, outliers)
- Dealing with more complex data types (record data, heterogeneous data, bag of words), data transformations, feature engineering
- Dimension reduction: heuristic, manual, etc.
- Volume reduction: sampling etc.
Course Optimization
INFM121ML.b
Lecture
Prof. Dr.-Ing. Astrid Laubenheimer
German
2/2
60 hours in total, including 30 hours of contact study.
Module exam
Course Optimization Exercise
INFM122ML
Exercise
Prof. Dr.-Ing. Astrid Laubenheimer
German
3/2
90 hours in total, including 30 hours of contact study.
Hands-on Work 1 Semester (graded)
Faculty of Computer Science and Business Informatics, Department of Computer Science