Recognition of human activities based on decision optimization model
Xueying Yang, Gang Huang · 2021 4th International Conference on Algorithms, Computing and Artificial Intelligence · 2021
Human activity recognition is a technology that uses a variety of sensors to identify and classify its activity data. Human activity recognition is widely used in the fields of medical treatment, daily health care and elderly care. Human activities are characterized by complexity and diversity, and it is difficult to clearly divide actions. The current human identification system is still extremely challenging. Therefore, we propose a decision optimization recognition model, which improves the recognition accuracy through model decision selection. The decision optimization recognition model adopts four typical statistical learning algorithms to construct a human activity recognition model. Algorithms include K-nearest neighbor algorithm (KNN), Bayesian algorithm (NB), support vector machine algorithm (SVM), and random forest algorithm (RF). The classification accuracy is used as the performance index of the algorithm to judge the recognition accuracy. The experimental results show that the recognition accuracy rate of human activities through the decision-making optimization recognition model is as high as 96.8%. Among the four typical statistical learning algorithms, the random forest algorithm has the best performance.