Research On The Application Of Predicting The Score Of College Entrance Examination Based On Content Matching

When applying for the college entrance examination,the score line of the college entrance examination plays a vital role.At present,there are few researches on the prediction of college entrance examination scores in our country,there are still gaps in the research on admission scores of specific majors,and the literature on practical applications is relatively lacking,so this article is very valuable for the research on college entrance examination scores.Due to the large amount of data in the college entrance examination and too many characteristic influence values,this paper selects the admission score data of the majors of the excellent planning classes in universities as the data support,and realizes the prediction of the minimum scores of the majors through model construction.The main research work of this paper is as follows:Firstly,through the webpage to obtain college professional admission data,and the existing admission data for heterogeneous data source integration.For similar repeated record data,the improved edit distance algorithm is used to calculate the similarity,and the adjacent record sorting algorithm is used to merge.For the missing data of the highest score,the mean difference interpolation method is used to fill in,and high-quality sample data is obtained through data cleaning.Secondly,before training the model,standardize the sample data according to the differences between the original data of college admissions.Use labeling coding to encode the classification features,and select 9features such as school,major,province,discipline,batch,average score,highest score,provincial control line,and year as input features,and use the lowest score as the input feature Output characteristics.In order to build a predictive model with higher accuracy,the data is normalized,and the sorted sample data is divided into a training set and a test set.Finally,1% and 2% error bands are preset,and the minimum score prediction model for colleges and universities is constructed based on BP neural network and support vector machine algorithm respectively.When constructing the BP neural network prediction model,trial and error method is used to determine the number of hidden layers and the number of neurons.When constructing the support vector machine prediction model,the grid search method and cross-validation are used to determine the optimal parameter set.Through experimental analysis,it is obtained that the prediction errors of the two experimental models meet the preset error band of 2%.When the preset error band is 1%,the prediction accuracy of the support vector machine is higher than that of the BP neural network,indicating that the network model based on the support vector machine algorithm achieves higher prediction accuracy and stronger generalization ability of the lowest score line of colleges and universities.Have better prediction results.