Academic Journal of Computing & Information Science, 2023, 6(6); doi: 10.25236/AJCIS.2023.060618.
Jingkai Zhang1, Tiandao Luo2
1School of Information Engineering, Zhengzhou Technology and Business University, Zhengzhou, 451400, China
2School of Electrical and Control Engineering, Henan Urban Construction University, Pingdingshan, 467000, China
At present, human action recognition technology has been widely used in intelligent monitoring, fatigue driving warning, fall detection, family rehabilitation training and other fields. In order to accurately identify human actions, this paper uses a variety of machine learning models. At the same time, in order to improve the accuracy of recognition, this paper uses a variety of feature data sets to train the model. Through experiments, it is found that the model trained by the feature data set after PCA dimensionality reduction has the best comprehensive effect. The prediction accuracy of logistic regression algorithm, KNN algorithm and LightGBM algorithm has been significantly improved. Compared with the models trained by other feature data sets, the recognition accuracy has been improved by 6% -20%, reaching 0.89, 0.87 and 0.83 respectively.
Human Behavior Recognition, Machine Learning, PCA, Grid Search, LightGBM
Jingkai Zhang, Tiandao Luo. Human behavior recognition based on machine learning. Academic Journal of Computing & Information Science (2023), Vol. 6, Issue 6: 115-120. https://doi.org/10.25236/AJCIS.2023.060618.
[1] Huang Tao. Research on human motion recognition technology based on MEMS inertial sensors [D]. Harbin University of Commerce, 2021.
[2] Bao Yanyan. Research and application of machine learning in pose recognition [D]. Xi'an University of Architecture and Technology, 2018.
[3] Altun K, Barshan B, Tunçel O. Comparative study on classifying human activities with miniature inertial and magnetic sensors [J]. Pattern Recognition, 2010, 43(10): 3605-3620.
[4] Han Songshan. Research on human motion recognition based on multi-sensor joint [D]. University of Electronic Science and Technology of China, 2018.
[5] Zhuo S, Sherlock L, Dobbie G, et al. Real-time Smartphone Activity Classification Using Inertial Sensors—Recognition of Scrolling, Typing, and Watching Videos While Sitting or Walking[J]. Sensors, 2020, 20(3): 655.
[6] Jian Ding, Sun Yue. Multi-classification prediction of flight delays based on LightGBM [J]. Journal of Nanjing University of Aeronautics and Astronautics, 2021, 53(6): 847-854.
[7] Jin M , Zhang J , Huang T , et al. Research on Human Action Recognition Based on Global-Local Features of Video[C]// International Conference on Pattern Recognition and Machine Learning. IEEE, 2021.