| Summary: | 碩士 === 輔仁大學 === 企業管理學系管理學碩士班 === 102 === In financial industry, the environment is getting much harsher than ever before. To be outstanding in financial sector, bankers have been tried to satisfy the need of customers as they can whether in service quality or in service area. Considering the characteristics of the consumer credit loans which are less risky and thriving, bankers are keen to sell consumer credit loans. The objective of the proposed study is to explore the performance of classification model for classifying fiduciary purchasing behavior using ensemble learning techniques. This study proposed a hybrid Logistic regression, Discriminant analysis, Extreme learning machine, Support vector machine and Artificial neural networks method to upgrade the performance of classification model comparing to single learner above.
To demonstrate the effectiveness of the ensemble learning approach, classification tasks are performed on one consumer dataset of credit loans telemarketing. 15 variables are adopted in this study. The result shows that the proposed approach is better than other five single classification models.
|
|---|
