PCA-SVM stock selection model based on value investment
Xiaofeng Hui · Xi'an Gongcheng Keji Xueyuan xuebao · 2009
In order to accommodate the value regression trends of the stock market and guide rational investment,under the guidance of the value investment theory,the stock option model based on value investment is set up by PCA-SVM method.In accordance with the characteristics set of the shares value,the principal component analysis is used to realize the extraction of the characteristics of the shares value.Then multiple sets of different training samples are adopted to train support vector machines and tested inside and outside the samples if they were success.The Shanghai′s 180 index component stocks are identified by this model.The finding is that the PCA-SVM stock option model has a good stock picking ability.