A Study on Caregiver Activity Recognition for the Elderly at Home Based on the XGBoost Model

Zhonghua Liu,Shuang Zhang,Huihui Zhang,Xiuxiu Li
DOI: https://doi.org/10.3390/math12111700
IF: 2.4
2024-05-31
Mathematics
Abstract:This paper aims to discuss the implementation of data analysis and information management for elderly nursing care from a data-driven perspective. It addresses the current challenges of in-home caregivers, providing a basis for decision making in analyzing nursing service content and evaluating job performance. The characteristics of caregivers' activities were analyzed during the design of a wearable device-wearing scheme and a sensor data collection system. XGBoost, SVM, and Random Forest models were used in the experiments, with the Cuckoo search algorithm employed to optimize the XGBoost model parameters. Based on the control group experiment, it was confirmed that the XGBoost model, after adjusting the parameters using the Cuckoo search algorithm, exhibited better recognition performance than the SVM and RandomForest models, and the accuracy reached 0.9438. Wearable devices present high recognition accuracy in caregiver activity recognition research, which greatly improves the inspection of caregivers' work and further promotes the completion of services. This study actively explores the applications of information technology and artificial intelligence theory to address practical problems and effectively promote the digitalization and intelligent development of the elderly nursing care industry.
mathematics
What problem does this paper attempt to address?
The main problem this paper attempts to address is the identification and management of caregiver activities in home care. Specifically, the study aims to analyze and manage caregiving activities in home care through a data-driven approach to improve the quality and efficiency of caregiving services. The core issues of the paper are as follows: 1. **Designing an appropriate wearable device scheme**: - One of the main challenges faced by the research is how to design a suitable wearable device scheme for Caregiver Activity Recognition (CAR). This includes how to effectively extract features and accurately distinguish different behavioral activities. 2. **Improving the accuracy of activity recognition**: - Traditional static models have low recognition accuracy due to individual differences when identifying behavioral activities. Therefore, the research needs to explore new methods to improve the accuracy of activity recognition. 3. **Dealing with complex real-world caregiving activities**: - Actual caregiving activities often involve complex actions that are easily confused, making accurate recognition difficult. The research needs to address this issue to ensure accurate identification of various caregiver activities. To address the above challenges, the study adopted the XGBoost model and optimized the model parameters through the Cuckoo Search algorithm. Experimental results show that the XGBoost model, after parameter adjustment, outperforms the SVM and Random Forest models in recognizing caregiver activities, with an accuracy rate of 0.9438. Additionally, the study explored the application of wearable devices in caregiver activity recognition, providing strong support for improving the quality of caregiving services.