A Comparison of the Efficiency of Algorithms and Feature Selection Methods for Predicting the Success of Personal Overseas Money Transfer
Keywords:
Decision Tree, Naive Bayes, K-Nearest NeighborsAbstract
The study's purpose was to compare the efficiency of algorithms and feature selection methods for predicting the success of personal overseas money transfers. The information of personal overseas money transfers certificate using in their study was 2016-2017 and 51,901 records. This research was used three classification techniques: Decision Tree, Naive Bayes, and K-Nearest Neighbors which compare the efficiency of algorithms the predicting techniques. The efficiency testing can be predicted using Cross Validation method by RapidMiner Studio 8 program. Then, we experimented to find out the result of the efficiency that has maximum accuracy from the study we found that using decision tree technique with attribute. The accuracy is 99.90%, K-Nearest Neighbors 99.55%, and Naive Bayes 96.71%, respectively. Comparison of results we are able to take the technique which has the most accuracy to predict the success of personal overseas money transfers.