Skip to main content
. 2020 Aug 10;20(16):4474. doi: 10.3390/s20164474

Figure 5.

Figure 5

A comparison between the performance of three methods in the detection of FOG episodes on the training set (n = 57); freezing index as a single feature, Samà et al. [23] feature set with an SVM classifier and the proposed feature set with the same SVM classifier. The training set was chosen for this comparison since it consists of a larger dataset. Similar results were obtained with the test set. * p < 0.05, values are presented as median.