Implementation of decision tree using C4.5 algorithm in decision making of loan application by debtor (Case study: Bank pasar of Yogyakarta Special Region)

A bank loan is very common in society. Before getting a loan, the applicant has to go through survey step done by loan analyst to evaluate whether the applicant is eligible or not to get the loan. A loan analyst must be very thorough in predicting if the applicant is qualified to get the loan to prevent repayment stoppage. Therefore, a tool is needed to support the loan analyst in decision making. Decision tree is a prediction model using tree structure or hierarchical structure. Decision tree is one of the most popular classification methods since it is easy to undertand. C4.5 is a decision tree algorithm commonly used to generate a decision tree since it has a high accuracy in decision making. C4.5 algorithm is the successor of ID3, in which the root and the parent are selected not only based on information gain but also on gain ratio as parent selection by finding the split information first. Dataset used in this research is as many as 1000 data in which 70% is approved and the other 30% is rejected. This report discusses the performance of C4.5 algorithm in identifying debtor's eligibility. The result shows that the biggest precision value is 78.08% reached by C4.5 algorithm with data partition of 90%:10%. The biggest recall value is 96.4% with data partition of 80%:20%. However, from the same training data result, ID3 reaches precision value of 71.51% and recal value of 92.09%. In this case, this result shows that C4.5 algorithm is proven to have a higher accuracy level and better than ID3.