Feature Selection for Improving the Usability of Classification Results of High-Dimensional Data

The semiconductor industry is one of the many fields of applications, where the usage of computational tools is inevitable in order to analyze huge amounts of data. As quality demands and production costs are rising, data mining methods move into the focus of interest. They should help maximizing the generation of information from the data being gathered during the production process and therewith supporting the human experts. In this paper the authors present a system being adopted in the evaluation of measurements from a semiconductor production process.