Decision tree-based credit decision support system

In this article, our aim is to demonstrate a new approach on financial credit decision support based on artificial intelligence, which, beside the classic methods, gets more and more prevalent in economics [1]. Our goal was to work out and achieve a decision tree-based credit decision method which is able to process and appraise data both in large numbers and one at a time. Beside the quick and robust solution, the efficiency of the algorithm was also an important aspect. Our improved algorithm of the traditional decision tree building is going to be demonstrated, and we will also touch upon the incidental problems that have occurred during the development, and their solutions. The efficiency of the algorithm in different situations is also going to be demonstrated with tests, and the results of the algorithm are going to be compared with a bank system´s results on the same set of data.