Table 2.
Mean average error (MAE) of the linear regression model predicting the probability of the correct class for each worker using the same metric data and resulting classifier weights for the workers and videos used in Table 1 and mean probability of the correct class calculations for a distinct set of videos for a distinct set of workers.
| Input Features | MAE (%) All data points |
|---|---|
| MSCL1, MPIL1, mean time | 10.93 |
| MSCL1, mean time | 13.03 |
| MPIL1, mean time | 11.50 |
| MSCL1, MPIL1 | 10.41 |
| MSCL1 | 11.87 |
| MPIL1 | 10.91 |
| Mean time* | 12.10 |
*
Mean time as the only feature is the only configuration of input features that requires a different set of data points: N=102 instead of a subset of size N=62 for all other configurations.