三峡大学学报（自然科学版）2018，Vol.40Issue(3)：88-90,112,4.DOI:10.13393/j.cnki.issn.1672-948X.2018.03.019
基于模糊信息粒化与支持向量机的上证50ETF量化择时研究
Research on Quantitative Timing of SSE 50ETF Based on Fuzzy Information Granulation and Support Vector Machine
摘要
Abstract
The subject matter of SSE 50ETF stock index options is SSE 50ETF.So,quantitative analysis of 50ETF is conducive to guide the investment of 50ETF and 50ETF options;and then earning good returns.By this,in this paper,the algorithm of quantitative timing of SSE 50ETF based on the support vector machine (SVM)and the fuzzy information is proposed;the algorithm is according to the historical data of SSE 50ETF,used the opening price,the closing price,the highest price etc.of a certain period of time as the input information of SVM,carried on Fuzzy information granulation of the next 5 days'closing price,and used the rise and fall of granulation as the training set label of SVM,then implemented labeled training on the system. Finally,the trained system is used to predict the future ups and downs of 50ETF.Compared among the pre-diction algorithm combining SVM with fuzzy information granulation,SVM algorithm and the partial particle swarm optimization algorithm incorporating SVM.Experimental results show that using the prediction algo-rithm combining SVM with fuzzy information granulation,under the local market,the accuracy rate has in-creased by more than 15 %;and under the global market the correct rate has increased by more than 16 %;thus the timing performance is improved.So,the quantitative timing strategy of this paper has practical guid-ance for 50ETF and 50ETF option investment;and it has certain application value.关键词
上证50ETF/期权/量化择时/模糊信息粒化/支持向量机(SVM)Key words
SSE 50ETF/option/quantitative timing/fuzzy information granulation/support vector machine(SVM)分类
信息技术与安全科学引用本文复制引用
伍呈呈,夏平,雷帮军,胡蓉..基于模糊信息粒化与支持向量机的上证50ETF量化择时研究[J].三峡大学学报（自然科学版）,2018,40(3):88-90,112,4.基金项目
国家自然科学基金(联合基金)重点项目(U1401252) （联合基金）
国家自然科学基金项目(61272237) （61272237）
省重点实验室开放基金项目(2015KLA05) （2015KLA05）
