Data mining na reálných datech

The aim of this bachelor's thesis is to find interesting relationships in a set of real data by using techniques of data mining. This is achieved by searching for answers to given analytical questions. The analysis was based on a dataset describing the housing market in Russia. The data were obtained from Kaggle.com, which is a platform focused on data processing. The whole work is divided into theoretical and practical part.The theoretical part summarizes available basic information about data mining and process of knowledge discovery in databases (KDD), which data mining is a part of. The definitions, history of this discipline and the most used methodologies, especially the CRISP-DM methodology, which determines the steps according to which the practical part is performed, are introduced to the reader. Furthermore, the GUHA method, original Czech method for exploratory data analysis, is presented. Its procedures are implemented by the LISp-Miner system, which is used for analysis in the practical part of the thesis. Its description is also part of the theoretical part.The practical part describes the whole process of DZD on selected real data step by step - from data selection to analytical procedures and interpretation of the results. The academic tool LISp-Miner was used for working with the data. The CF-Miner and KL-Miner were chosen as modules to answer the formulated analytical questions. These modules implement the GUHA procedures of the same name. In conclusion, the obtained results are interpreted, and the whole analysis is summarized, including the extent to which the objectives were met and problems encountered during the work.