Skip to main content
. 2022 Dec 16;22(24):9903. doi: 10.3390/s22249903

Table 5.

A comparison of this work with state-of-the-art models’ work for activity detection. PYP (Postural Yaw Perturbation), GT (Gait Task), TTHP (Toe Tapping with Heel Pin), CT (Circling Test), WLE (Walk-like events), LDA (Linear Discriminant Analysis), MLP (Multilayer Perceptron), 1D-CNN (One-dimensional Convolution Neural Network).

Author Objective Type of Task ML Algorithm Accuracy
This work Classification of PD from HC PYP kNNf 96%
Aich et al. [23] Detection of FoG GT SVM 88%
Caramia et al. [25] Classification of PD from HC GT SVM 80%
Klucken et al. [40] Classification of PD from HC GT, TTHP, CT LDA, AdaBoost, SVM 81%
Naghavi et al. [41] Detection of FoG GT kNN, SVM, MLP 97%
Atri et al. [42] Classification of PD from HC WLE 1D-CNNs 90%