Data Mining Industrial Applications

Novel, advanced sensors, dynamic development of information technologies as well as modern high-performance computers applied in different fields of human activity result in large amount of data. Consequently, these data, grouped in the data sets are both large and complex. The complexity come from the several mutually excluding factors like acquisition with different sensors at various times, frequencies or resolutions. The increasing size and complexity of data in different practical, often industrial branches stands the challenging problem for nowadays scientific disciplines. Regarding to above facts, it is notable that these techniques progressively displace many traditional methods (eg. visualization and statistics) that are no longer suitable for the analysis. Like schematic drawings and mathematical equations were formerly necessary to obtain competitive advantage, currently data mining techniques place similar role. They enable to make scientific discoveries, gain fundamental insights into considered physical process and advance in their better understanding. Although “data mining” term reflects somehow its idea of mining the data in general, it had a varied origins and history, evolving during time and borrowing and enhancing ideas of different fields. These domains have included statistics, image processing, machine learning, mathematical optimization, information retrieval etc. That’s why data mining has multidisciplinary nature (Kamath, C., 2009). It is worth to point out that in some disciplines like statistics, terms “data mining” or “data dredging” have negative connotation, regarding to the fact they were used to describe extensive searches through data. Statisticians tend to ignore the developments in data mining (Duebel, C., 2003), (Tang, Z., 2005). The term “data mining” originally referred to a single stage of Knowledge Database Discovery (KDD) process. While KDD is a nontrivial process of identifying valid, useful and understandable patterns in data, data mining means which patterns are extracted and enumerated from data. This idea combines regularities finding (hidden for human) with computer's calculation speed in large amount of data (Jacobson, R. & Misner, S., 2005). Some practitioners (Simoudis, E., 1996) refer to data mining as the process of extracting valid, previously unknown, comprehensible and actionable information form database and using it to make crucial business decisions. This approach joins data mining with data warehouse and divide the process into four actions carry out on data: selection, transformation, mining and results interpretation. However circular definition considers it as process of extracting useful information from data. Some data miners preserve distance form terminology debate, focusing on how the