Multiple costs based decision making with back-propagation neural networks |
Abstract:
The current research investigates a single cost for cost-sensitive neural networks (CNN) for decision making. Thismay not be feasible for real cost-sensitive decisions which involve multiple costs. We propose to modify theexisting model, the traditional back-propagation neural networks (TNN), by extending the back-propagationerror equation for multiple cost decisions. In this multiple-cost extension, all costs are normalized to be in thesame interval (i.e. between 0 and 1) as the error estimation generated in the TNN. A comparative analysis of accuracydependent on three outcomes for constant costs was performed: (1) TNN and CNNwith one constant costCNN-1C), (2) TNN and CNN with two constant costs (CNN-2C), and (3) CNN-1C and CNN-2C. A similar analysisfor accuracywas also made for non-constant costs; (1) TNN and CNNwith one non-constant cost (CNN-1NC), (2)TNN and CNN with twnon-constant costs (CNN-2NC), and (3) CNN-1NC and CNN-2NC. Furthermore,we comparedthe misclassification cost for CNNs for both constant and non-constant costs (CNN-1C vs. CNN-2C andCNN-1NC vs. CNN-2NC). Our findings demonstrate that there is a competitive behavior between the accuracyand misclassification cost in the proposed CNN model. To obtain a higher accuracy and lower misclassificationcost, our results suggest merging all constant cost matrices into one constant cost matrix for decision making.Formultiple non-constant costmatrices, our results suggestmaintaining separate matrices to enhance the accuracyand reduce the misclassification cost. |
| Keywords: |
Cost-sensitive Neural networks Multiple costs Misclassification |
| Author(s): |
Guang-Zhi Ma , Enmin Song , Chih-Cheng Hung, Li Su , Dong-Shan Huang |
| Source: |
Decision Support Systems 52 (2012) 657–663 |
| Subject: |
تصمیم گیری |
| Category: |
مقاله مجله |
| Release Date: |
2012 |
| No of Pages: |
7 |
| Price(Tomans): |
0 |
|
بر اساس شرایط و ضوابط ارسال مقاله در سایت مدیر، این مطلب توسط یکی از نویسندگان ارسال گردیده است. در صورت مشاهده هرگونه تخلف، با تکمیل فرم گزارش تخلف حقوق مؤلفین مراتب را جهت پیگیری اطلاع دهید.
|
|
|