基于SVM的高校录取分数预测模型

Prediction model for college admission score based on SVM

由于随机性大,影响因素多,导致高校录取分数预测困难,准确度不高,相关研究较少.针对这种情况,尝试基于支持向量机(SVM)理论,结合我国高考录取模式,建立SVM模型对高校录取分数进行初步预测.通过对预测结果的定量分析,证明预测效果较理想,预测的平均绝对误差为7.6分,同时验证了SVM预测模型在高校录取分数预测中的可行性.

The college admission score prediction is with the high difficulty, the low accuracy and the less related research, because of its high randomicity and multiple influencing factors. Aiming at the situation, based on the support vector machine（ SVM ）theory, the prediction model for college admission score is constructed to carry out the preliminary prediction, which is combined with the admission mode of Chinese college entrance examination in the paper. By the quantitative analysis of the prediction results, the ideal prediction effect is achieved, the mean absolute prediction error is 7.6, and the feasibility of the SVM prediction model is verified for college admission score prediction.