| The college admission score is a critical basis for filling out college applications.Being admitted to a satisfactory college is a dream goal for candidates and parents.Therefore,establishing a reasonable model to predict the admission score is crucial for improving the accuracy of enrollment.The new college entrance examination policy implements a “Choose 3 of 6 Subjects” model,which weakens the reference value of the division of arts and sciences data and reduces the effectiveness of existing research on predicting college admission score lines.In response to the initial implementation of the new college entrance examination policy,there is less amount of data,and a lack of consistency and comparability between the ”Choose 3 of 6Subjects” model and the division of arts and sciences model data.This paper establishes a prediction model for college and professional threshold scores,providing an effective method for filling out college applications.In response to the lack of consistency and comparability in data between the“Choose 3 of 6 Subjects” model and the division of arts and sciences model of the new college entrance examination,this paper first uses the window moving average method to process the frequency data of the “one point one segment table” under the two models.It is found that the distribution of science subject data was closer to the “Choose 3 of 6 Subjects” model data distribution than the distribution of arts subject data,and passed the k-s test.Then,this paper applies the standard normalization method to the data in the “ one point one segment table ” from 2018 to 2022 and establishes a score conversion model,effectively solving the problem of the lack of comparability of data from different years.In order to predict the 2022 admission threshold scores of 630 undergraduate colleges enrolling students in Shandong Province,this paper first converted the 2018 to 2021 score lines of each college into equivalent scores in 2022 through a score conversion model,and then used the GM(1,1)model to predict the 2022 admission score lines of each college.The analysis of the prediction results showed that the average absolute error of 630 colleges was 6.29 points,and the average absolute error of 76 colleges was 16.99 points.The reason for the large prediction absolute error is that there are significant fluctuations in the admission score line of colleges,resulting in significant differences in conversion scores.Therefore,for colleges with a difference in conversion scores greater than 15 points for two consecutive years,the range optimization model is applied to optimize the conversion scores.After optimization,the average absolute error of the predicted threshold scores for 76 colleges was 7.92 points,a decrease of 9.07 points compared to the previous one.The average absolute error of the predicted score line for 630 colleges was 5.2 points,a decrease of 1.09 points.In order to improve the accuracy of college applications,the preliminary screening of colleges is conducted using candidate scores and college prediction score lines.Based on six evaluation indicators that can reflect the comprehensive level of colleges,this paper applied the K-means clustering algorithm of paper swarm optimization to cluster and screen colleges,and proposed a “motivation,maintenance,and stability”discrimination method.Applying this method to identify the majors of the selected colleges,the results of identifying the majors of colleges can help candidates to fill out college applications.This paper solved the problem of lack of consistency and comparability of data from different years,and established a prediction model for college threshold scores.According to the comprehensive level of the colleges and the factors of the candidates themselves,it provides effective methods of filling out college applications,offering practical plans and theoretical basis for candidates to accurately fill in their college applications. |
PDF Full Text Request