Research Article
Lung Cancer Classification Employing Proposed Real Coded Genetic Algorithm Based Radial Basis Function Neural Network Classifier
Table 8
Comparison of the network training parameters for the LIDC dataset.
| Parametric measures | Hybrid Logistic Regression-Artificial Neural Network Approach [9] | Hopfield Neural Network & Fuzzy Clustering Approach [10] | Back Propagation Neural Network Approach [11] | SRGWO-ELM Approach | Proposed RCGA-RBFNN Approach |
| Norm | 378.67 | 231.98 | 196.72 | 121.47 | 109.54 | MSE Error | 0.9231 | 0.9745 | 0.1120 | 0.0097 | 0.0089 | Training Efficiency % Mean | 86.25 | 81.26 | 94.57 | 98.83 | 99.25 | Training Efficiency % STD | 6.71 | 6.47 | 5.64 | 6.05 | 5.96 | Testing Efficiency % Mean | 90.35 | 88.61 | 95.84 | 97.12 | 98.96 | Testing Efficiency % STD | 6.13 | 6.09 | 5.45 | 6.03 | 6.01 | Hidden neurons | 19 | 7 | 12 | 5 | 3 | Accuracy % | 92.8 | 89.07 | 97.14 | 98.64 | 99.03 | Time min | 3.27 | 2.96 | 3.5 | 2.01 | 1.76 |
|
|