Search for: [Abstrakt = "Data mining techniques are largely based on machine learning algorithms. They are to serve to extract data models which, due to their large information content, are not recognized by people. Data redundancy poses a problem both for data mining algorithms as well as people, which is why various methods are used in order to reduce the amount of analyzed data, including data mining methods such as feature selection. The article outlines basic issues linked with feature selection and contains an analysis of five feature selection algorithms belonging to the filter category. Results obtained by each method were validated with the help of CART decision tree algorithms. The CART analysis revealed that the results of each of the five algorithms are acceptable"]
