A verification study on the effect evaluation of curriculum ideological and political education using an improved Kalman filter algorithm combined with the back propagation neural network
The implementation of ideological and political education in courses in colleges and universities reflects a novel,multi-dimensional,and multi-category concept of educating students.Its core purpose is to enhance moral education,foster talents and promote collaborative education.The evaluation and verification of teaching effect is an important stage in the process of college reformation of teaching.Since the ideological and political teaching process involves variations in students'emotions,attitudes,thoughts and other aspects,the quantitative evaluation and verification of its teaching effect has been one of the persistent problems to be solved urgently in the implementation process of ideological and political education in college courses for long time.Therefore,in order to verify,feedback,and optimize the evaluation of ideological and political teaching effects,a relatively succinct,open,and iteratively updated quantitative empirical formula was constructed based on the teaching practice of ideological and political education in courses.At the same time,the score of the empirical formula for the implementation effect of ideological and political education is accompanied with some potential factors such as degree of difficulty for an examination paper and teaching quality to construct the state equation and observation equation of the Kalman filter(KF)model for student performance prediction.Furthermore,taking advantage of the significant advantages of the back propagation neural network(BPNN)in the analysis of complex nonlinear relationships,a Kalman filter algorithm model combined with BPNN was constructed to verify a typical nonlinear system of student performance prediction.In practice,the powerful self-learning ability of BP neural network was used to optimize the state parameters in the Kalman filter model.The ideological and political education effect of each step in teaching design was analyzed through the measurement innovation of student performance prediction,and the results was used to reversely verify,adjust,and optimize the evaluation factors and weights of the empirical model formula for ideological and political education effect.Using the implementation of ideological and political education in college physics curriculum as a case study to predict the scores of this courses for recent four years by Matlab.The results showed that the Kalman state transition matrix parameters obtained through neural network training and identification were better compared with the traditional KF algorithm,and the learning process was more convergent.It is evident that the prediction ability processed by the Kalman filter algorithm was significantly improved.In addition,the results indicated that the impact of ideological and political education in courses on students'studies was positively correlated with the scores.Students'various performance indicators were also steadily ameliorated,and there were significant improvements in aspects such as recognition,participation,and evaluation of courses and classroom teaching.This model provides a basis for the evaluation of the teaching effectiveness of ideological and political education in courses and its improvement.Thus,it is important to construct and improve an effective quantitative evaluation model for the teaching effect of ideological and political education in courses that conforms to the characteristics of various disciplines and majors.That can effectively promote the construction and development of ideological and political education in college courses,and it also contributes to the achievement of knowledge objectives and ability objectives in course teaching,and further ensures the continuous improvement of educational quality in higher education institutions.
effect evaluationvalidationKalman filterback propagation neural networkideological and political education
国家科技期刊平台
NSTL
万方数据
