Data mining algorithms which are used during data analysis have more and more important role nowadays in comprehending and explaining data sets, which are hard to interpret for the human mind. This problem usually caused by the complexity of the data, or more commonly the large quantity of it. The efficiency of the analysis is based on finding the correct algorithm, with the best parameters which is an iterative process and takes most of the time of the analysis. Understanding the operation of algorithms which are used during the analysis is usually different from interpreting the concrete model which generated for the utilized data set.
This thesis demonstrates common data mining algorithms, and widely used visualization techniques, and methods to apply them together, in order to get a feedback about the created models behavior, and this way it improve the efficiency and the speed of data analysis. Furthermore, the paper demonstrate the application of these methods on two different data sets.