ICA applications in data mining
Data mining has received a lot of attention in academic research as well as commercial applications. With the growing size and complexity of data, the standard techniques have started showing their limitations in obtaining meaningful results and discovering new patterns from conventional data analysis. Several new methods have been devised to analyze today's enormously huge and inherently complex databases. Also many intelligent computing technologies such as neural networks and genetic algorithms have shown promising results for data mining and knowledge discovery. This thesis introduces independent component analysis (ICA), a well-known technique for signal processing and blind source separation, as a potentially better alternative to traditional data mining techniques. Different kinds of data are analyzed using ICA. Results of data analysis with ICA are compared to results of traditional rule-type data mining tools.