Although the film and television media sector is seeing significant growth, there is still a lag in film and television media majors at colleges and institutions. As a result, the colleges must keep pace with the current development trend. In order to better prepare students for the film and television industry, film and television media majors must rethink their teaching methods and promote students’ artistic ideas and professional skills. In the context of surging numbers of students and limited teaching resources, traditional teaching has highlighted many problems. Some film and television media majors have set up corresponding computer courses to make up for the deficiencies of traditional teaching through computer multimedia technology and network technology. This work is oriented to the design of computer courses for film and television media majors. First, this work aims at the current situation of computer courses in film and television media majors, analyzes the existing problems and causes, and explores ways to solve them. Second, this work proposes the IGA-BP model based on neural network to evaluate the teaching quality of computer courses for film and television media majors. In view of the excellent global optimization ability of GA and the defects of the BP algorithm itself, this work adopts the improved GA algorithm to optimize the BP network, and establishes an IGA-BP network combination model with higher prediction accuracy. Third, this work has carried out sufficient experiments, and the experimental results have verified that the IGA-BP network can effectively evaluate the teaching quality of computer courses in film and television media majors. In addition, the comparative experiment also verifies that the computer course design scheme proposed in this work can effectively improve the teaching quality.