数据挖掘在高校招生预测中的应用研究

高校招生人数准确预测问题,高校招生人数由于受到当前国家政策、社会需求、社会经济状态等因素综合影响,使招生人数变化存在非线性、复杂性。传统预测线性模型难以进行准确预测,预测准确率低。为了提高高校招生人数预测准确率,提出一种数据挖掘的高校招生人数预测模型。首先采用数据重构方法,获取多维高校历史数据,然后采用主成分分析对其进行处理,消除数据之间的重复信息,最后采用非线性预测能力强的支持向量机进行建模。采用某高校1994-2010年招生数据对模型的性能进行仿真,预测准确率高,证明建立的模型可以为高校招生未来人数预测提供参考。

Research on Data Mining for College Enrolment Prediction

Under current national policy,social demand,social economic status and other factors,the student enrollment changes nonlinearly and is of complexity,and it is difficult for traditional prediction model to accurately predict and the forecast accurate rate is low.In order to improve the university enrollment forecast accuracy,the paper proposed a forecast model of university enrollment based on data mining.First,the method of reconstructing data was used to obtain multi-dimensional university history data.Then,the principal component analysis was used to treat the data to eliminate the data duplication among information.Finally,the nonlinear prediction ability of the support vector machine was used in modelling.The enrollment data from 1994 to 2010 were used for simulation experiment.The prediction accuracy rate is very high,and the model can be used for predicting future enrollment of College enrollment.