Table 1.
Prediction performance of the machine learning-based methods in terms of mean absolute error for each of the motility values and the overall average.
| Classical Machine Learning Results | ||||
|---|---|---|---|---|
| Method | Progressive | Non-progressive | Immotile | Average Mean Absolute Error |
| Baseline | ||||
| ZeroR | 17.260 | 7.860 | 13.660 | 12.927 |
| Participant Data Only | ||||
| Elastic Net | 15.198 | 9.525 | 13.441 | 12.721 |
| Gaussian Process | 15.556 | 9.762 | 13.474 | 12.931 |
| Simple Linear Regression | 15.416 | 9.281 | 13.601 | 12.766 |
| SMOreg | 15.355 | 9.441 | 12.959 | 12.585 |
| Random Forests | 13.312 | 8.886 | 11.905 | 11.368 |
| Random Tree | 17.801 | 10.952 | 14.984 | 14.579 |
| Tamura Image Features Only | ||||
| Elastic Net | 14.400 | 7.750 | 12.190 | 11.447 |
| Gaussian Process | 13.230 | 7.260 | 11.920 | 10.803 |
| Simple Linear Regression | 13.520 | 8.170 | 12.690 | 11.460 |
| SMOreg | 13.220 | 7.260 | 11.920 | 10.800 |
| Random Forests | 13.530 | 7.400 | 12.060 | 10.997 |
| Random Tree | 18.700 | 9.960 | 16.520 | 15.060 |
| Tamura Image Features and Participant Data | ||||
| Elastic Net | 14.130 | 9.890 | 11.750 | 11.923 |
| Gaussian Process | 13.700 | 10.120 | 11.460 | 11.760 |
| Simple Linear Regression | 13.940 | 10.240 | 11.410 | 11.863 |
| SMOreg | 13.710 | 10.140 | 11.460 | 11.770 |
| Random Forests | 13.510 | 10.000 | 11.340 | 11.617 |
| Random Tree | 18.660 | 13.270 | 16.960 | 16.297 |
The best performing algorithm in each category is in bold.