Parameters | Training-Validation-Testing (70% - 15% - 15%) | Training-Validation-Testing (50% - 25% - 25%) | ||||
SCG | LM | RPROP | SCG | LM | RPROP | |
Number of Neurons (Input) | 6 | 6 | 6 | 6 | 6 | 6 |
Number of Neurons (Hidden) | 10 | 10 | 10 | 10 | 10 | 10 |
Number of Neurons (Output) | 2 | 2 | 2 | 2 | 2 | 2 |
Learning Coefficient | 0.75 | 0.75 | 0.75 | 0.75 | 0.75 | 0.75 |
% of Error | 1.90% | 1.30% | 3.20% | 1.30% | 8.40% | 2.60% |
Number of Epochs | 14 | 11 | 12 | 17 | 8 | 14 |
Training* | SCG | LM | RP | SCG | LM | NRP |
Gradient | 0.0064 | 0.0022 | 0.043 | 0.007 | 0.0037 | 0.0416 |
Training Optimization Time (s) | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
Number of Training Dataset (Fall-Non Fall) | 109 (36 - 73) | 109 (36 - 73) | 109 (36 - 73) | 77 (28 - 49) | 77 (28 - 49) | 77 (28 - 49) |
Number of Validation Dataset (Fall-Non Fall) | 23 (7 - 16) | 23 (7 - 16) | 23 (7 - 16) | 39 (12 - 27) | 39 (12 - 27) | 39 (12 - 27) |
Number of Test Dataset (Fall-Non Fall) | 23 (7 - 16) | 23 (7 - 16) | 23 (7 - 16) | 39 (10 - 29) | 39 (10 - 29) | 39 (10 - 29) |
Number of Misclassification (Training) | 0 | 0 | 2 | 0 | 7 | 0 |
Training Dataset CDR | 100.00% | 100.00% | 98.20% | 100.00% | 90.90% | 100.00% |
Number of Misclassification (Validation) | 3 | 2 | 2 | 2 | 2 | 2 |
Validation Dataset CDR | 87.00% | 91.30% | 91.30% | 94.90% | 94.90% | 94.90% |
Number of Misclassification (Test) | 0 | 0 | 1 | 0 | 4 | 2 |
Test Dataset CDR | 100.00% | 100.00% | 95.70% | 100.00% | 89.70% | 94.90% |
Overall CDR | 98.10% | 98.70% | 96.80% | 98.70% | 91.60% | 97.40% |
Specificity | 100.0% | 100.0% | 98.1% | 100.0% | 87.6% | 98.1% |
Sensitivity | 94.0% | 96.0% | 94.0% | 96.0% | 100.0% | 96.0% |