Abstract
In the rapid urbanization process in China, due to reasons such as employment, education, and family reunification, the number of mobile population without registered residence in the local area has increased significantly. By 2020, the group had a population of 276 million, accounting for over 20% of the total population, making significant contributions to urban economic development and resource optimization. However, the health status of migrant populations is affected by unique issues such as occupational risks and socio-economic disparities, which play an important role in personal welfare, social stability, and sustainable economic growth. The deterioration of the health of the floating population will lead to a decrease in productivity, an increase in medical expenses, and an increase in pressure on the public health system. In order to analyze and predict the main elements affecting the well-being of transient population, this study uses advanced machine learning algorithms such as principal component analysis, backpropagation (BP) neural networks, community analysis, random forest models, etc. Principal component analysis will identify and extract the most important variables that affect the health status of mobile populations. The BP neural network models the nonlinear interaction between health determinants and health outcomes. Community analysis divides the floating population into different health records and promotes targeted intervention measures. The random forest model improves the accuracy and universality of predictions. The insights generated by these models will help develop health policies and intervention policies to improve the health status of mobile populations, narrow disparities, and promote social and economic stability. Integrating data-driven methods and emphasizing a shift towards correct, effective, and impactful public health management provides a robust framework for understanding and addressing the complex health issues faced by mobile populations.