Credit risk estimations based on some data of applicants were carried out in RapidMiner software platform. For this aim a Decision Tree Model and Logistic Regression Model were built and compared. The model processes included the data mining and analysis, section of independent variable subset, splitting the data into the training and test subsets, building the models on the train subset and applying it to the test one, the performances (confusion matrices) calculations and summarizing the prediction ability of the models.