This paper takes the students of neurology specialty in a medical school as the research object and analyzes the current situation of its educational team construction. A machine learning algorithm based on weighted plain Bayes is used to construct an evaluation model, through which different weights are given to each index to explore the effectiveness enhancement strategy of neurology ideology and politics teaching team construction. Finally, the teaching performance and comprehensive ability of Civics and Politics were compared between the two groups of students to explore the effect of infiltrating the course Civics and Politics in neurology teaching. The results showed that 54.65% believed that teaching neuroscience ideology and politics in team construction should be improved as soon as possible. The mean value of “exerting the guiding role of professional teachers of Civics and Politics” was 23.76 points, which was also the furthest away from a perfect score, and was the focus of the subsequent team-building concern. The teaching performance and comprehensive ability of the students in the experimental group were better than that of the reference group (P<0.001), and the full penetration of the relevant elements of the course Civics in the teaching of clinical internships can transform the students’ ideological concepts and conscious behaviors.