Research On Improvement Of RBF Neural Networks And Its Application To Course Teaching Effects Evaluation

With China’s twelfth five-year development MSC and much starker choices-and graver consequences-in the beginning of the development plan, the economic development of our country enters the new normal, the quality of higher education in the new challenge. Teaching is still the core mission of higher education, improve the teaching quality is the key to promote the sustainable development of the reform, nowadays in universities for improving the quality of teaching, especially the newly built undergraduate course colleges and universities for dislocation development with old public colleges and universities, actively seeking transformation development and change. And the success of teaching reform, to see whether its advantages and disadvantages and effects on promoting teaching quality improvement. Improve the teaching quality and strengthen teaching management must be through the establishment of scientific teaching quality evaluation system. The university teaching quality of the basic support of the survival and development, seeking international competition of higher education also need to constantly improve the quality to increase core competitiveness. The teaching quality of ascension in colleges and universities to improve the quality of teaching units should be enhanced to implement, is embodied in every is to pay special attention to the course construction of professional quality, and the most fundamental point is the quality of the teachers teach every course.In order to improve the fairness and objectivity of the evaluation of the teaching effect of the course, first, a new evaluation model based on improved RBF neural networks is studied in this paper. By dynamically changing the control parameters of the L-M algorithm, the model can improve the convergence ability of RBF networks. Then, on the basis of comprehensive analysis of all the factors on the evaluation of the teaching effect of the course, a set of index system including two levels of 25 indicators is established. The simulation results verify the rationality of these indexes. Finally, the validity of the proposed model is verified by the real history data of the teaching evaluation in a private college. The establishment of the model provides a new way to evaluate the teaching effect of the course objectively and fairly.