基于CNN-BiLSTM-ATT混合模型的高校高考录取分数预测研究OA

The prediction of college entrance examination admission scores in Colleges and Universities is of great significance to candidates,parents,and educational institutions.However,this prediction work is challenging due to the influence of multiple factors such as examination difficulty,enrollment strategies of Colleges and Universities,and the scale of candidates.Therefore,this paper proposes a hybrid model based on CNN-BiLSTM-ATT to predict the admission scores of Colleges and Universities.The model first uses Convolutional Neural Network(CNN)to extract the local features of admission scores in Colleges and Universities,then learns the long-term dependency in the time series through Bidirectional Long-Short Term Memory(BiLSTM),and finally introduces the Attention Mechanism(ATT)to enhance the focus on key year data to improve prediction performance.The experimental results show that the CNN-BiLSTM-ATT model has high accuracy and generalization ability in predicting the admission scores of Colleges and Universities.Compared with other models,it more effectively captures the changing trend of admission scores and achieves better evaluation indices,providing a valuable reference for college entrance examination volunteer application.