clustering

The objective of the current investigation was to conduct a biomechanical analysis of human gait based on the Unsupervised machine learning – Artificial Intelligence approach. Twenty-eight junior active males participated in the study. Following the placement of the markers, the participants were asked to complete the gait task in a 10-meter gateway where the dominant leg contact was placed on the third step and non- non-dominant leg on the fourth step. The task was executed in two separate attempts, first by the preferred speed of the participants and second with a steady speed of 100BPM. The Hierarchical approach consisting of Nearest Neighbor and the utilization of Z score was employed to discern uniform gait biomechanical patterns of the entire participant according to the values of joint angles and joint moments in both conditions - preferred and steady speeds by SPSS software version 26 (p<0.05). Considering a combination of both kinematics and kinetics parameters, in preferred speed, the hip and knee in the vertical direction for both dominant and non-dominant limbs are classified in one cluster, but in a steady speed, the hip in mediolateral direction and knee in the vertical direction for both dominant and non-dominant limbs are presented in one cluster. The kinematic and kinetic variables are useful in gate clustering to categorize gait patterns. These variables can be subdivided into homogeneous subgroups for a more detailed understanding of human locomotion.