K-means-XGBoost Integrated Classification Method and Its Application in Sleep Data Analysis

In the context of big data, traditional classification methods face challenges in processing complex high-dimensional data. This study proposes a classification method that combines Kmeans clustering and XGBoost. The elbow method is used to determine the optimal number of K-means clusters. After integrating the cluster labels into the original data, XGBoost is used to build a classification model. The experimental results show that the optimal number of clusters for this data set is $\mathbf{4}$. After 5fold cross-validation, the XGBoost model has an average score of 0.9259 and an accuracy of 0.9075. It can accurately identify the four categories of light sleep type, deep sleep type, irregular sleep type and regular sleep type obtained by clustering. The study shows that this combined method shows good performance in the sleep data classification task, providing a certain method reference and practical basis for subsequent research in related fields.