A Study of Mother-Infant Behavioral Relationships Based on Support Vector Machine Predictive Models

This paper focuses on an exploratory study of the relationship between mothers' physical and psychological conditions and infants' behavioral characteristics and sleep quality. The study was conducted through data preprocessing, including outlier exclusion and missing value treatment, to ensure the accuracy of the data. First, the patterns of the effects of mothers' physical and psychological indicators on infants' characteristics were identified through box-and-line graph visualization and scatterplot analysis; second, Pearson correlation coefficients were introduced to further confirm the relevance of these effects. Next, a support vector machine (SVM) model was used to classify and predict infant behavioral characteristics. In addition, optimization models were constructed to minimize the treatment costs of CBTS, EPDS, and HADS. Finally, an entropy weighting method was used to comprehensively evaluate infant sleep quality, and a multiple linear regression model was constructed to predict infant sleep quality, providing new perspectives and methods for future research.