基于公告分类校准的多指标SVM创业板选股方案设计

发布时间：2018-02-02 17:09

本文关键词： 创业板支持向量机股东人数公告股票　出处：《上海师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：我国民众对于金融市场投资的需求大大提高,人们的投资需求和优质的投资去向之间存在极大地矛盾,资金多流向基于供求力量的股票二级市场。同时投资者自身经验匮乏且缺乏金融知识,急于找寻能够在股票市场获得超额收益的方法,所以对于股票价值走势规律的研究自然成了热门课题。本文以创业板市场股票为研究对象,基于我国证券投资市场的特色和证券估值的本质,提出具体的假设条件,考虑技术分析,基本面分析以及特色指标筹码集中度情况,利用SVM(支持向量机)模型来根据设计的策略即输入指标将股票资产分类,选择准确率高的核函数筛选优质资产池;再根据市场对利好利空消息的短期效应将上市公司的公告信息分类,影响经营情况的且无违纪和潜在风险的、只是改变财务鼓励政策的以及有违规和潜在风险的,保留优质资产池。构建一种符合投资者习惯兼之拥有金融逻辑且收益效果不错的选股策略方案。
[Abstract]:The demand for investment in the financial market is greatly increased in our country, and there is a great contradiction between the demand for investment and the destination of high-quality investment. At the same time, investors lack experience and lack of financial knowledge, so they are eager to find ways to obtain excess returns in the stock market. Therefore, the research on the trend of stock value has naturally become a hot topic. This paper takes the gem stock market as the research object, based on the characteristics of China's securities investment market and the nature of securities valuation. Put forward the specific hypothetical condition, consider the technical analysis, the fundamental analysis and the characteristic index chip concentration degree situation. The SVM (support Vector Machine) model is used to classify the stock assets according to the designed strategy, that is, the input index, and the kernel function with high accuracy is selected to select the high quality asset pool. Then according to the short-term effect of the market on the good news, the public announcement information of the listed company is classified, which affects the operating situation and has no violation of discipline and potential risk. Just change the financial incentive policy and the violation and potential risk, retain the pool of high-quality assets. Build a stock selection strategy that is consistent with investors' habits and has financial logic and good returns.
【学位授予单位】：上海师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F832.51

【参考文献】