Research Article
Predicting Tooth Surface Loss Using Genetic Algorithms-Optimized Artificial Neural Networks
Table 2
Actual TSL scores and ANN predictions for those scores of the test (number = 15) subjects.
| Subject number | TSL score | Net prediction | Difference (residual) | Accurate prediction rounded to nearest integer | Accurate predictions rounded to nearest 5 | Accurate predictions when rounded to nearest 8.3 |
| 81 | 13 | 12.01 | 0.99 | | Yes | Yes | 82 | 16 | 11.9 | 4.1 | | Yes | Yes | 83 | 11 | 7.4 | 3.6 | | Yes | Yes | 84 | 5 | 4.2 | 0.8 | | Yes | Yes | 85 | 9 | 15.4 | −6.4 | | | Yes | 86 | 9 | 17.3 | −8.3 | | | Yes | 87 | 16 | 20.3 | −4.3 | | Yes | Yes | 88 | 9 | 16.2 | −7.2 | | | Yes | 89 | 14 | 14.49 | −0.49 | Yes | Yes | Yes | 90 | 16 | 14.49 | 1.51 | | Yes | Yes | 91 | 9 | 14 | −5 | | Yes | Yes | 92 | 12 | 12.3 | −0.3 | Yes | Yes | Yes | 93 | 9 | 15.8 | −6.8 | | | Yes | 94 | 17 | 21.9 | −4.9 | | Yes | Yes | 95 | 11 | 15.2 | −4.2 | | Yes | Yes |
| Overall accuracy of predictions | 13.33% | 73.33% | 100% |
|
|