BP_Adaboost算法的改进及在首轮融资时总票房分类预测中的应用

为获得改进的分类算法BP_Adaboost,利用思维进化算法(MEA)和列文伯格-马夸尔特算法(LM)结合改进的BP神经网络作为弱分类器,由改进的弱分类器集成得到MEA-LM-BP_Adaboost算法.提出了基于MEA-LM-BP_Adaboost算法的首轮融资时总票房分类预测方法,该方法包括变量选取及操作化处理、网络参数优化、MEA改进弱分类器、LM算法改进弱分类器、MEA-LM-BP_Adaboost算法的流程设计、待预测电影验证6个部分.选用2013～2018年的245部国产电影作为样本验证该预测方法和模型,测试集分类准确率可达73.3％.最后在模型准确率、稳定性、K折交叉验证3方面进行模型整体性能比较,结果表明本文提出的模型整体性能最好.

The Improvement of BP_Adaboost Algorithm and Its Application to the Total Box Office Classification Prediction for the First Round of Financing

To obtain the improved BP_Adaboost algorithm for classification, the BP neural network improved by the combination of the mind evolution algorithm(MEA), and the LM algorithm was used as the weak classifier, and then obltaned the MEA-LM-BP_Adaboost algorithm through integration of the improved weak classifier. A total box office classification prediction method based on MEA-LM-BP_Adaboost algorithm was proposed. The method includes variable selection and operation processing, network parameter optimization, weak classifier improvement by MEA, weak classifier improvement by LM algorithm, MEA-LM-BP_Adaboost algorithm''s flow design and movie prediction. 245 domestic movies released during 2013~2017 were taken as the samples to validate the method. The classification accuracy of the test set was 73.3%. Finally, the overall performance of the model was compared in terms of model accuracy, stability, and K-fold cross-validation. The results show that the proposed model has the best overall performance.