中文 | English

J. Hefei Univ. Tech. (Nat. Sci.)

合肥工业大学学报(自然科学版)

JOURNAL OF HEFEI UNIVERSITY OF TECHNOLOGY (NATURAL SCIENCE)

Competent Authorities: Ministry of Education of the People's Republic of China

Sponsor by: Hefei University of Technology

ISSN 1000-1646, CN 34-1001/N

A method for stock turning point prediction with hidden semi-Markov model based on sentiment vector

Download Full Text

基于情绪向量的隐半马尔可夫模型股市拐点预测方法

Journal Information

J. Hefei Univ. Tech. (Nat. Sci.), Oct.2024, Vol.47 No.10: 1335-1340

DOI: 10.3969/j.issn.1003-5060.2024.10.006

Authors

YAO Hongliang, JIANG Yongsheng, YANG Jing, YU Kui

(School of Computer Science and Information Engineering, Hefei University of Technology, Hefei 230601, China)

Abstract and Keywords

Abstract: The sentimental stock market is the main reason for the high degree of uncertainty in the market trend. The stock trend prediction method using historical data directly is difficult to adapt to the variability of market sentiment, and the effect is not ideal in practical application. Aiming at the problem that it is difficult to predict the turning point of stock market due to the instability of market sentiment, a hidden semi-Markov model stock turning point prediction method based on sentiment vector(SV-HSMM) is proposed. Firstly, for market sentiment is not observable, the main features related to market sentiment are selected and fused into market sentiment by Markov blanket. Secondly, the HSMM is used to model the market environment, and the structural relations among market sentiment, market state and state duration are constructed. Furthermore, sentiment vector is introduced to smooth the variability of sentiment, and Kullback-Leibler(KL) distance is used to quantify sentiment heat. Finally, the dynamic inference of HSMM is used to predict the turning point of stock market. Experimental results show that the sentiment vector method has better prediction effect.

Keywords: market sentiment; sentiment vector; hidden semi-Markov model (HSMM); Kullback-Leibler (KL) distance

作者信息

姚宏亮，江永生，杨静，俞奎

（合肥工业大学计算机与信息学院，安徽合肥230601）

摘要和关键词

摘要: 股市的情绪化倾向是股票市场具有高度不确定性的主要原因, 直接利用历史数据的股票趋势预测方法难以适应市场情绪的多变性, 在实际应用中效果不理想。文章针对市场情绪的不稳定性导致股市拐点难以预测的问题, 提出一种基于情绪向量的隐半马尔可夫模型股市拐点预测方法(hidden semi-Markov model stock turning point prediction method based on sentiment vector, SV-HSMM)。针对市场情绪不可观察性, 选取与市场情绪相关的主要特征, 使用马尔可夫毯融合成市场情绪; 利用隐半马尔可夫模型建模市场环境, 构建市场情绪、市场状态和状态持续时间之间的结构关系; 引入情绪向量平滑情绪的多变性, 并利用 Kullback-Leibler(KL) 距离量化情绪热度; 利用隐半马尔可夫模型的动态推理实现股市拐点预测。结果表明情绪向量方法具有更好的预测效果。

关键词: 市场情绪；情绪向量；隐半马尔可夫模型（HSMM）；Kullback-Leibler（KL）距离

Funding

国家重点研发计划资助项目（2020AAA0106100）；国家自然科学基金资助项目（62187620；62176082）

Previous Next

Back to Contents