Verifikace pravděpodobnostních rozdělení chí-kvadrát testem a její účinnost v závislosti na způsobu rozdělení dat do tříd

Abstract

This bachelor thesis is primarily focused on verifying selected probability distribution by chi-square test and by examining the most appropriate chi-square form of the test depending on how the data is divided into classes. The bachelor thesis describes the most important statistical methods in quality management which must comply certain prerequisites for proper operation. Each the statistical method has the assumption of a specific probability distribution of data which are the input for that method. Probability distribution should be verified by the statistical test. One test is a chi-square test. The chi-square test needs to divide the data sample into different classes, which are different rules. The aim of this bachelor thesis is evaluation which variants of this test, depending on how data is divided into classes gives the most reliable results.