Table 3. Ranking of the vocal features based on their mutual information with UPDRS level discretized according to the determined optimal threshold that can be discriminated by machine learning methods.
| Ranking | Dysphonia Measurement | MI Score |
|---|---|---|
| 1 | DFA | 0.0413 |
| 2 | PPE | 0.0302 |
| 3 | RPDE | 0.0287 |
| 4 | HNR | 0.0277 |
| 5 | NHR | 0.0202 |
| 6 | Jitter:PPQ5 | 0.0189 |
| 7 | Jitter(%) | 0.0189 |
| 8 | Jitter(Abs) | 0.0163 |
| 9 | Shimmer:APQ11 | 0.0163 |
| 10 | Jitter:DDP | 0.0155 |
| 11 | Jitter:RAP | 0.0154 |
| 12 | Shimmer(dB) | 0.0120 |
| 13 | Shimmer | 0.0112 |
| 14 | Shimmer:APQ3 | 0.0084 |
| 15 | Shimmer:DDA | 0.0084 |
| 16 | Shimmer:APQ5 | 0.0081 |