Artificial intelligence and intelligent algorithms provide data-driven methods for research on pension scheme selection. To address the issue of unclear mechanisms underlying the influence of perceived service quality, this study proposes a hybrid analysis model combining logistic regression and random forest. Using Python, the researchers performed data cleaning, standardization, variable encoding, and the division of training and testing sets. Building upon the traditional logistic regression model’s identification of the direction of influence of explanatory variables, the study introduced random forest to identify nonlinear relationships and key factors.The experimental results show that the Random Forest model achieved an accuracy of 0.859 and an AUC of 0.901, both higher than the 0.813 and 0.846 recorded by the Logistic Regression model, indicating that intelligent algorithms possess superior predictive performance. Service convenience, the level of intelligent services, and service reliability were identified as the primary influencing factors.
