Table 2 The neural network solution based on the training data set and testing (holdout) data set classified at 95.5% and 73.7% accuracy respectively for skill level (beginner, intermediate and advanced) for the motor fitness tests.
|
Predicted |
|
|
1 |
2 |
3 |
% Correct |
| Training |
|
|
|
|
| 1.Beginner |
8 |
0 |
1 |
88.9% |
| 2.Intermediate |
0 |
5 |
0 |
100.0% |
| 3.Advance |
0 |
0 |
8 |
100.0% |
| Overall Percent |
36.4% |
22.7% |
40.9% |
95.5% |
| Testing |
|
|
|
|
| 1.Beginner |
9 |
1 |
0 |
90.0% |
| 2.Intermediate |
3 |
2 |
1 |
50.0% |
| 3.Advance |
0 |
0 |
3 |
100.0% |
| Overall Percent |
63.2% |
15.8% |
21.1% |
73.7% |
