Gait Analysis through Machine Learning Algorithms

Technology combined with medical science has been an irreplaceable achievement of modern times. The 27 experimental subjects and 16 control groups were classified in four unique classes by classification model made by using MATLAB tools. The diseased as well as healthy subjects were classified according to features derived from the clinically obtained Gait Neurodegenerative (GaitNDD) dataset from the physio net platform. A classification model has been constructed. The multiclass classification models gave the characteristics in a single graph for all the four unique classes. The four classes were Amyotrophic Lateral Sclerosis (ALS), Huntington's, Parkinson's, and Control. The characteristics obtained were compared with the theoretical results available. Here, the validation accuracy obtained was 98.8% for the SVM model and 52% for the Nearest Neighbour model for the considered dataset.