PENERAPAN CLASSIFICATION BASE ON MULTIPLE ASSOCIANTON RULE PADA ANALISA RESIKO KREDIT USAHA MIKRO DI BANK SYARIAH MANDIRI KCP PRAYA
ABSTRACT: Based on data of
customer Micro Credit in BSM KCP Praya, there is a fact customer with troubled
credit status is high enough percentage level. Of the total 171 customers,
found as many as 42 clients with problematic status or by 24.56%. This is
despite the presentation is still below the current credit status, it still
would be a problem at KCP Praya because it would result in the loss and affect
the balance of assets or general praya KCP. For the analysis of micro credit in
KCP Praya has not used methods or techniques specific, but in another study
that examines in particular the problem of credit ratings many use
classification algorithm C4.5, SVM, neural network, logistic regression and
classification methods traditional or other common. Particularly in this study,
the authors use the technique Classification Asociative approach CMAR, because
it is known proven Accuracy better than traditional methods. Results of
measurement accuracy for 171 datasets micro credit customers in BSM KCP Praya
obtained value reached 99.42% accuracy. Credit risk analysis that will be
examined starting from the identification of variables that influence the
existing loan in the training data, generate a model Classification Association
Rule (CARs) with CMAR method, do ranking and pruning rule to get the best
classifier models, and carry out testing of data to predict credit risk. At
last evaluate performance and accuracy of the results in the form of credit
risk prediction
Keywords: Accuracy,
Association classification (AC), Classification Association rule (CARs), CMAR, Credit
risk Analysis
Penulis: Moh. Farid Wajdi,
Hamzan Ahmadi, L. M. Samsu
Kode Jurnal: jptinformatikadd170280
