Home im.ve Group im.ve Lab Projects Publications Teaching News, Events, Activities Contact us News

DataMining and Visualization Project in Cooperation with HITeC/Plath

In this project, a study compared various datamining techniques to evaluate their suitability for feature extraction in radio emission data. Goal of the study was to propose a solution for automatic feature recognition of important events in the data. Examples are clustering techniques, independent component analysis, model-based event recognition, and spatio-temporal techniques. Visualization was included as data-mining method as well as to show results. In the current second phase, a prototype to implement and test the proposed method on real data is under construction. im.ve as as cooperation partner in HiTeC implements and evaluates the non-supervised data-mining techniques and the visualization of intermediate and final results.

 

contact: Matthias Haringer

People involved:

Haringer