2.3.2. Field Study

For the evaluation of the 3.2-km field study dataset, we used the MAE to evaluate the total distance of the runs. After segmenting the strides and calculating the stride lengths for each stride, we accumulated the single stride lengths and compared them to the ground truth value of 3200 m. The reason for choosing the MAE for this evaluation was the fact that the absolute deviation of the ground truth value is of great importance to runners. For the *Acceleration* and *Deep Learning* algorithms, we computed the regression models based on the lab study dataset. Due to having different subjects participating in the lab and the field study, the results were not overfitted. The GPS measurements of the total distance of the individual runs were also evaluated by comparing them to the gold standard value of 3.2 km.
