Data Mining for technical applications
From data to knowledge
Data Mining refers to the process of automatically obtaining valid, novel, potentially useful and understandable knowledge from large amounts of data (According to Fayyad, Piatetsky-Shapiro, Smyth and Uthurusamy 1996). Techniques from different fields such as statistics, machine learning, pattern recognition etc. are used. Applications of data mining can be found in areas such as marketing, medicine, biochemistry, real-time systems and many others. In technical applications, knowledge is often derived from sensor signals, that is, from numerical data. These are often "insecure", eg. For example, readings may be inaccurate or even missing.
Content of the lecture
The lecture "Data Mining for Technical Applications" introduces general fundamentals in the field of data mining. Among other things, the data mining process, data preprocessing and basic algorithms for clustering and classification are discussed. In the second part of the lecture, the focus is on some classifiers that are functionally very similar, but nevertheless originate from completely different worlds: Radial basis function networks in the field of neural networks, support vector machines in the field of statistical learning theory and probabilistic classifiers, which are based on probabilistic considerations. all three belong to the current state of the art in the field of data mining and have very different properties depending on the type of application. Finally, the lecture deals with the combination of different techniques in the form of ensembles.
Content of the exercise
The aim of the exercise is that the participants can independently solve thasks in the field of data mining. Practical computer exercises with the freely available tool RapidMiner therefore take up a lot of space.
The conclusion of the lecture is a competition in which the participants can independently apply the acquired knowledge.
- Formal, genral information about the lecture (eg assignment to areas of application, credits, examination type) can be found in the module handbook. Degree Programm -> Examination Regulations -> Module
- Up-to-date information about the current lecture (eg slides, bibliographical references, examination dates) will be provided in the Moodle of the University of Kassel.