發(fā)布時間：2018-01-02 10:26

  本文關鍵詞：基于理性指標的馬爾可夫鏈股市態(tài)勢預測方法　出處：《計算機工程與應用》2017年22期 　論文類型：期刊論文

  更多相關文章： 理性指標 馬爾可夫鏈 特征背離 主成分分析

【摘要】：由于股票市場存在人為擾動性,使得基于情緒的股市預測算法效果不佳。針對股市的誘多誘空問題,提出一種基于理性指標的馬爾可夫鏈股市態(tài)勢預測算法(RI_MCA)。提取股市的主要理性特征,并對這些理性特征進行量化;通過主成分分析將這些理性特征融合成理性指標,并利用理性指標獲取股市的買賣點;將買賣點所對應的股市狀態(tài)引入到馬爾可夫鏈中,實現(xiàn)股市態(tài)勢預測。在理性指標和股市狀態(tài)相背離情況下會降低買賣點的可靠性,因而通過將特征背離引入到RI_MCA算法中提出了RICD_MCA算法,RICD_MCA算法根據(jù)特征背離程度對RI_MCA算法的結果進行調整優(yōu)化。在上證指數(shù)上的實驗比較與分析結果表明,RICD_MCA算法具有更高的預測精度。
[Abstract]:Due to the existence of the stock market disturbance, the stock market sentiment prediction algorithm based on the induced effect of poor Youkong stock market, put forward a prediction algorithm of Markov chain based on rational stock market trend index (RI_MCA). The main reason for feature extraction of the stock market, and to quantify these rational characteristics; through principal component analysis, the rational features into a rational and rational use of indicators, indicators of access to stock market trading; trading at the stock market will be introduced to the state in the Markov chain, prediction of stock market trend in the rational index and the stock market deviates from the state will lower reliability point of sale, so the characteristic deviation is introduced into RI_MCA algorithm is put forward according to the results of the RICD_MCA algorithm, RI_MCA algorithm on the deviation from the characteristics of the adjustment and optimization of the RICD_MCA algorithm. Compared with the Shanghai index on the experiment The analysis results show that the RICD_MCA algorithm has a higher prediction accuracy.

【作者單位】： 合肥工業(yè)大學計算機與信息學院;
【基金】：國家自然科學基金(No.61175051,No.61175033) 國家重點基礎研究發(fā)展規(guī)劃(973)(No.2013CB329604)
【分類號】：F832.51;O211.62
【正文快照】： 1引言股票市場是一個不穩(wěn)定的非線性動態(tài)變化的復雜系統(tǒng)[1],有效地預測股市態(tài)勢是一個公開的難題[2]。同時,由于股市存在人為擾動現(xiàn)象,如誘多和誘空,使得股市的有效預測更加困難。技術指標在股市態(tài)勢預測中起到了重要的作用,普通股民和投資機構都會考慮技術因素;因而,研究者也，

本文編號：1368819