Non-parametric probabilistic neural network: statistics summary of 100 repetitions
| Samples | C1 | C2 | C3 | C4 | C5 | C6 |
|---|---|---|---|---|---|---|
| Training sample | ||||||
| Mean correct classification | 98.91 | 92.45 | 87.30 | 80.95 | 74.42 | 59.50 |
| Standard deviation | 3.00 | 5.89 | 6.45 | 3.63 | 9.73 | 11.67 |
| Coeff variation (%) | 3.04 | 6.37 | 7.39 | 4.48 | 13.08 | 19.61 |
| Minimum | 84.61 | 73.33 | 72.41 | 70.69 | 47.62 | 37.04 |
| Maximum | 100.00 | 100.00 | 100.00 | 86.21 | 95.00 | 77.78 |
| Range | 15.38 | 26.67 | 27.58 | 15.52 | 47.38 | 40.74 |
| Conf interv upper | 99.76 | 94.12 | 89.13 | 81.98 | 77.19 | 62.82 |
| Conf interv low | 98.061 | 90.78 | 85.46 | 79.92 | 71.66 | 56.18 |
| Validation sample | ||||||
| Mean correct classification | 89.20 | 89.03 | 75.21 | 86.60 | 79.41 | 58.67 |
| Standard deviation | 6.41 | 6.42 | 6.41 | 6.67 | 3.49 | 10.47 |
| Coeff variation (%) | 7.19 | 8.30 | 12.99 | 7.70 | 4.39 | 17.80 |
| Minimum | 72.73 | 73.20 | 40.06 | 69.00 | 72.22 | 38.46 |
| Maximum | 100.00 | 100.00 | 94.12 | 96.70 | 86.79 | 78.57 |
| Range | 27.27 | 26.80 | 54.04 | 27.70 | 14.57 | 40.11 |
| Conf interv upper | 91.03 | 91.52 | 75.51 | 88.50 | 80.41 | 61.65 |
| Conf interv low | 87.38 | 89.02 | 67.15 | 84.70 | 78.43 | 59.60 |
Stability estimates and confidence intervals of the classification function were tested by 100 bootstrap repetitions with random selection of the replaced samples. The average correct classification was higher, and the function stable for C1, C2, C3, and C4, while C6 had large standard deviations and coefficients of variation during both training and validation. C: cluster