Module 39-M-Inf-ABDA_a Advanced Big Data Analytics / Big Data Machine Learning

Non-official translation of the module descriptions. Only the German version is legally binding.

The emphasis of this module is on understanding and application of machine learning and artificial intelligence techniques for the analysis of large datasets. Examples of such techniques often deal with popular internet based applications, such as recommender systems or web advertisements, in any case techniques that are influential for modern daily life.
The basic motivation is to develop a deeper understanding for the algorithms, the statistics that underlie such techniques and the processes they drive. The goal is that seminar participants feel well equipped, and poised to actively engage in related processes, both in theory and in professional practice.
Depending on concrete choice of topics, the seminar/project can start with tutorials run by the teacher, to introduce seminar participants to the topics of choice. Presentations or projects then are supposed to practice how to understand, how to interpret and evaluate original literature, or to implement relevant theory in practical applications.
While in case of a presentation, the focus is on presentation techniques, programming techniques are stressed in case of programming projects. Drafting a scientific report is supposed to generate text that complies with actual scientific and ethical standards; an important aspect is the exact reproducibility of the facts and processes discussed.

Contents of tutorials and literature relevant for the seminar cover topics where the employment of artificial intelligence and machine learning for the analysis of big datasets plays a prevalent/pivotal role. Popular, ubiquitous examples are the identification of communities, or the prediction of (not yet recognizable, because hidden) links in social networks, the employment of deep neural networks in predicting the fit of advertisements with (e.g. search engine) users, or the recommendation of products in online stores that promise maximal profits. Classification systems that evaluate big data streams, which need to be regularly updated, are equally relevant.

Participation in the lecture ‘Big Data Analytics’ is helpful. Basic knowledge of algorithms, data structures and artificial intelligence is required.

—