| Summary: | Mandiri Syariah Bank Branch Office of Luwuk, receives a very large
number of proposal credit in every month and needs a quick response. Thus, the
system should be developed to perform data mining in the data heap to be used for
specific purpose, one of the purpose is to analyze the risk of credit allowance.
Naive bayes classifier is an approach that refers to the bayes theorem,
which combine the prior knowledge and the new knowledge. So that, this
classifier is one of a simple classification algorithm but has a high accuracy. this
research will prove the ability of naive bayes classifier to classify the debitur data
that contains information of credit allowance in Mandiri Syariah Bank Branch
Office of Luwuk. Before doing the classification, data of debitur needs to pass a
preprocessing method. Then the classification process by naive bayes classifier
was done after passing the preprocessing method. After the data is classified, it
produces the probability of classification model to predict the class of next
debitur.
From the testing result, the program shows the smallest value of the
accuracy is 80% by using 100 records of sample and generating highest accuracy
for about 98,66% by using 463 records of sample. The testing results by Rapid
Miner 5.3 software obtained the smallest value of the accuracy is 64,79% by using
100 records of sample and the highest accuracy is 80,06% by using 463 records of
sample for naive bayesian classification. For the method of support vector
machine obtained the smallest value is 63,99% accuracy by using 100 records of
sample and the highest accuracy of 78,64% by using 463 records of sample.
|
|---|