Použití Business Intelligence pro analýzu malých podniků

Abstract

The subject of the submitted thesis is the topic of business analyses with a special emphasis on data mining techniques. The thesis covers procedures of performing these analyses, describes basic methods, and assesses the possibilities of their application in commercial praxis. The largest scope is given to the CRISP-DM methodology, especially the phases of preprocessing data, processing data and evaluating results. The practical part then applies the described methods to commercial data from a real company operating vending machines at schools, mainly using Rapidminer and MS SQL Business Intelligence Development Studio tools. It demonstrates the usage of a few basic data mining methods (Principal Component Analysis, Correlations, Associations, Clustering methods, Decision Tree). The obtained results are interpreted with respect to acquiring knowledge that could be commercially useful for the client.