Selecting of Pet Adopters using C4.5 Decision Tree Model Algorithm

Abstract

This study aims to select pet adopters based from the model created from the C4.5 decision tree algorithm to address the proper way for adoption of a pet through online examination. This study uses the method of the C4.5 algorithm by finding the highest gain value to create a decision tree model and generate a decision rules that will apply to a website for decision support if the adopter is qualified or not qualified to adopt. A confusion matrix is used to calculate the accuracy of the C4.5 decision tree algorithm. The information in the confusion matrix is needed to calculate the classification model's performance. This can be done by using performance metrics for the C4.5 algorithm that is based on accuracy, sensitivity, and specificity. The total number of datasets is 1800 which consists of 15 attributes, 1500 for training which consists of 100 instances, and 300 for testing the accuracy is consist of 20 instances. The results on testing the accuracy of the C4.5 decision tree algorithm using confusion matrix is 85%, sensitivity is 84.6% and specificity is 100%. Based on the result of testing the accuracy of the algorithm using the 20 dataset that results with the accuracy value of 85%, sensitivity is 84.6% and specificity is 100%. These results prove that generated model using C4.5 decision tree algorithm is efficient in selecting of pet adopters.