PENERAPAN ALGORITMA DECISION TREE C4.5 UNTUK PENILAIAN RUMAH TINGGAL

Abstract

There is still a possibility of assessment error homes as a reference value of credit, which will open opportunities for NPL. So we need a way of assessment (predictive value) is quite proportional, credible and accurate. Inaccurate predictions led to the planning of improper credit management. Prediction value of collateral house has attracted the interest of many researchers because of its importance both in theoretical andempirical.

Namely C4.5 decision tree algorithm, CART and CHAID that can be used for credit risk status. The third tree algorithm produces different models for the same data set. Therefore, this study aims to implement a C4.5 decision tree algorithm for the assessment of the residence. Evaluation results will be processing using precision and recall, and then compared and analyzed the results between assessors using other analysis methods (Naive Bayes, K-NN) with the results predicted by the method of classification algorithm C4.5. From here will look the accuracy of the implementation of C4.5.

Keywords :  Classification Algorithm, Decision Tree, C4.5, Assessment