With the acceleration of urbanization, the process and content of commuters" travel decisions have become more complex, such as choosing which travel mode and how far in advance to leave, etc., and these decisions are often coupled and influenced by each other. How to mine the influencing factors of travel decisions and explore the coupling relationship between each travel decision, so as to provide a basis for transportation planning and management, is an important issue. This article carried out a bivariate synergistic analysis for two important decisions of commuters" travel - travel mode and departure time. The RP questionnaire was designed to obtain commuters" travel data, and based on the cleaned data, a unidimensional model was built for the two decision variables and compared with the optimal one, and the results show that linear regression performs better in the prediction of departure time, and random forest performs better in the prediction of travel mode. On this basis, a new method of collaborative modeling was proposed. The comparison reveals that the accuracy of the collaborative model is enhanced, which improves the shortcomings of the previous single-decision analysis.