Table 6.

Optimal hyper-parameters after smoothing tuning. Summary table of the optimal hyper-parameters for the a posteriori smoothing step obtained using the strategy outlined in Section 3 and the grids defined in Table 3.

Model	Minimal Distance $\mathit{\tau }$ between Consecutive Change Points	Proportion $\mathit{\eta }$ of Walking Time Points above Which a Segment Is Labeled as Walking Activity	Precision	Accuracy
Decision Tree	$2.20$ s	$30\%$	$0.88$	$0.84$
Radial Basis Function SVM	$2.00$ s	$30\%$	$0.78$	$0.74$
3-NN	$1.60$ s	$45\%$	$0.89$	$0.84$
Logistic Regression	$2.45$ s	$30\%$	$0.83$	$0.78$