| Since the end of the last century,influenced by China's higher education enrollment policy,the development of China's higher education has entered a fast lane,and the quality of personnel training has become increasingly obvious.At the same time,the quality requirements of college students are increasing year by year,and how to guarantee the quality of talents for the society from colleges.Then we need to have a quality evaluation method for college students to meet the needs of the society.There is no standard for the evaluation of college students' training quality.Through the research on existing evaluation methods and evaluation indicators,the existing evaluation indicators are not comprehensive,the index weighting method is subjective,and the evaluation method is inefficiency.The research work in this paper as follows:First,aiming at the unreasonable construction of indicators,according to the needs of all sectors of society and the development requirements of college students,the evaluation index system of college students' training quality is constructed.The indicator system includes three first-level indicators of knowledge,ability,and quality;14 secondary indicators and 40 observation points such as professional knowledge,innovative ability,and ideological quality.Second,the evaluation of ideological quality mostly adopts the method of teacher evaluation,and it is not possible to comprehensively evaluate the students' ideological quality.This paper studies data collection methods and quantitative methods,and obtains observation point data from the educational administration system,library management system,teaching review materials,and questionnaires,and quantifies the secondary indicators.The use of teacher evaluation and student mutual evaluation in the quantification of indicators such as ideological quality makes the index quantitative perspective more comprehensive.Thirdly,it analyzes the advantages and disadvantages of commonly used weighting methods.In view of the subjective power of subjective weighting method and the poor interpretation of objective weighting method,a combination weighting method combining subjective weighting method and objective weighting method is proposed.The analytic hierarchy process is used to obtain subjective weights based on the questionnaires of college teachers,enterprise personnel,and government workers.The entropy weight method is used to calculate the objective weights based on the collected data.The distance function was used to calculate the subjective weight coefficient of 0.56 and the objective weight coefficient of 0.44.The combined weights are not only well explained,but also the subjectivity of the weights.Fourth,traditional evaluation methods are inefficient and cumbersome.Through the research on the quality evaluation of college students' training,it is proposed that the use of machine learning to evaluate the quality of college students' training can effectively improve the evaluation efficiency and simplify the evaluation process.The training sample data is obtained by combining the combination weighting method and the analytic hierarchy process,and the machine learning algorithm is trained by using the sample to obtain the machine learning evaluation model.After the training is completed,the model can be easily applied to the evaluation,which can improve the efficiency and simplify the evaluation process.The SVM is integrated using AdaBoost,the SVM kernel function selects the Gaussian kernel function,and the diversity measure function is introduced to improve the AdaBoost-SVM.The algorithm is tested with Iris dataset.The results show that the accuracy of the improved AdaBoost-SVM model reaches 0.92 on small datasets.The accuracy of the improved datasets is higher than that of SVM and BP neural networks.The student sample data was used to train and test the AdaBoost-SVM.The accuracy of the model iteration was 0.91,and the evaluation method was simple and effective. |
PDF Full Text Request