|
An Investigation of Machine Learning Techniques for Loan Default Payments Prediction |
|---|---|
| รหัสดีโอไอ | |
| Creator | Waraporn Jirapanthong |
| Title | An Investigation of Machine Learning Techniques for Loan Default Payments Prediction |
| Contributor | Wilawan Inchamnam, Waraporn Jirapanthong |
| Publisher | The Association of Council of IT Deans (CITT) |
| Publication Year | 2566 |
| Journal Title | Journal of Information Science and Technology |
| Journal Vol. | 13 |
| Journal No. | 1 |
| Page no. | 38-44 |
| Keyword | Loan Default Payments, Imbalance Data, Machine Learning, Ensemble Techniques, Dimensionality Reduction |
| URL Website | https://tci-thaijo.org/index.php/JIST |
| Website title | Journal of Information Science and Technology |
| ISSN | 2651-1053 |
| Abstract | Inbanking business, loan default paymentsof individual customersarecounted asrisksthat result in the loss of the business.Thus,some assessmentmechanisms are needed to assessthe risksof individual customerswho apply for personal loan products. This paper presents an investigation of machine learning techniques to predict loan default payments based on individual customers information backgrounds. The paper emphasis on the ensemble techniques that mostly used in banking business. Besidestheensemble prediction models, the principal component analysis is also used forfurther investigation. The experimental resultsshowed that all prediction models providedacceptable prediction of non-defaulting payment class, but providedunacceptable prediction ofdefault payment class. Thatis because the imbalance nature of the data and the featuresusedare not specificenough for the prediction modelsto classifythe minor class from the major class. This paper acts as an initial study of the credit default payment analysis. |
