In recent years there has been an increased adoption of data science methods in numerous fields and industries. With usage of statistical analysis, machine learning, programming, but also domain expertise, it finds and identifies different patterns that allow experts to make certain predictions. It also allows experts to derive valuable insights from numerous datasets with high level of intricacy, complexity and size. In this paper we focus on how financial institutions utilize data science methods for various tasks. One of the main ways financial institutions make use of such methods is for determining credit score, which highlights good and bad candidates for loan. Credit scoring would be a numerical evaluation of an individual with goal to calculate the probability of concluding a contract, making a purchase decision, or in order to be able to measure the creditworthiness of a person that needs a loan. The primary aim of this paper is demonstrating the main characteristic of (German Credit Data) data set. By applying our analysis, we were able to determine whether the individual would be considered a good or a bad candidate for receiving credit. After gathering the number of correctly classified instances, an experiment was conducted to compare three different decision trees, and how reducing its size and the number of leaves would affect further correctly classified instances.