Gait cycle prediction model based on gait kinematic using machine learning technique for assistive rehabilitation device

The gait cycle prediction model is critical for controlling assistive rehabilitation equipment like orthosis. human has recently used statistical models, but the dynamic properties of physiology limit current approach. Current models need detailed kinematic and kinetic data body as input parameters, measuring them requires special instruments, making difficult to use in real-world applications. In our study, three separate machine learning algorithms were create a model: Gaussian process regression, support vector machine, decision tree. algorithm model's parameters are height, weight, hip knee angle, ground reaction force (GRF). For better prediction, produced enhanced by incorporating different sliding window data. best period was DT with (t−3), which had root mean square error 3.3018 R-squared (R-Value) 0.97. projection focused on angle GRF feasible solution devices during cycle.