Research Article
A Pruning Neural Network Model in Credit Classification Analysis
Table 10
Classification accuracy rates comparison between PNN and other algorithms obtained from the literatures of the Japanese credit dataset.
| Authors (published year) | Algorithms (train-to-test ratios) | Classification accuracy rate (%) |
| Yu et al. (2008) [53] | (10 CV) | 75.82 | ANN (10 CV) | 80.77 | SVM (10 CV) | 79.91 | Neuro-fuzzy hybrid (10 CV) | 77.91 | Fuzzy SVM hybrid (10 CV) | 83.94 |
| Tsai et al. (2014) [52] | MLP (10 CV) | 84.38 |
| Our method (2017) | PNN (50%-50%) | 85.54 | PNN (5 CV) | 85.23 | PNN (10 CV) | 85.27 |
|
|