This study systematically analyzed the impact of maternal health indicators on infant behavior and sleep quality through machine learning models. In the search for the significant relevance of maternal mental health indicators to infant behavior and sleep quality, this study adopted the Spearman correlation analysis method. Next, logistic regression and random forest models were compared for performance in prediction. As a result, it was shown that the random forest model outperformed the logistic regression model in all evaluation metrics: accuracy, precision, recall, F1 score, and AUC value. In terms of AUC value, the random forest model performed far beyond the logistic regression model. It showed that the random forest model has a strong advantage in processing complex and multidimensional data and can provide a reliable prediction tool for maternal and child health research. Through systematic experimental design and rigorous model evaluation, this study not only revealed the mechanism of the impact of maternal health on infant growth, but also provided scientific basis and technical support for further research and practical application in related fields.