摘要
人们致力于寻找各种有效方法来规避股票市场风险和获得高收益,为此提出了许多分析股票价格运行趋势的技术。由于股票市场极为复杂,影响因素很多,导致每种技术在实际应用中都存在一定的缺陷和不足。
本文在前人研究的基础上,采用粗糙集与神经网络相结合的方法来分析股票价格运行趋势,研究的主要内容如下:
(1)利用粗糙集对股票原始数据进行约简,并采用了一种基于量子计算与遗传算法相结合的属性约简方法。
(2)通过对各种神经网络模型优缺点的比较,确定选用RBF网络模型。
(3)通过Matlab仿真实验,详细地比较了自组织选取中心算法、有监督学习选取中心算法和正交最小二乘法选取中心算法三种学习算法应用在预测股价走势预测中的准确度。
(4)建立了基于粗糙集和RBF网络的股价运行趋势的分析模型。将股票原始数据,经粗糙集约简处理之后,选择一种约简结果作为网络输入向量;并将网络输出向量走势图的拐点分成六类进行分析。
(5)最后,本文选取了几个具有代表性的上市公司的数据验证了本模型的可信度和实用性。
So far, people have been looking for all kinds of effective methods to avoid the risk of stock market and to obtain higher returns from stocks, so many technologies for pretending the trend of stock market have been produced. For the complexity of the stock market, many technologies have exposed their shortcomings and insufficiency.
In this paper, the author bases on the previous studies and uses rough sets and neural network to predict the trend of stock price. The main contends are as follows:
(1)At first, obtain the original data from stock market using Da Zhihui Software, and then use rough sets to extract representative data from original data. In this process, a new algorithm based on quantum computing and genetic algorithm has been used and has some advantages compared with other algorithms.
(2)Considering the various advantages of RBF network, the author chooses it for establishing the pretending model.
(3)As there are many learning methods for RBF network, so choosing the effective method has been the focus. In this paper, the author compares four algorithms and chooses the best from them through experiments.
(4)At last, the author establishes a model based on the front research. The input vectors are choosed from the rough sets results; the output vector is divided six kind and drawned in the result chart to help users make decisions more effectively.
Finally, the author uses representative data to validate the correctness and credibility of the model .
引文
[1]刘德红.股票投资技术分析.经济管理出版社,2004:149～195.
[2]林俊国.股票投资学[M].北京:经济科技出版社,2006.
[3]邵道明.股市黄金搭档:股市技术指标最佳经典组合.经济管理出版社, 2009.9.
[4]周正庆.证券知识读本.中国金融出版社.1998.
[5]张文修,吴伟志等.粗糙集理论与方法[M].北京:科学出版社,2001.
[6] Z.Pawlak. Roughsets:theoretical aspects of reasoning about data [M].Kluwer Academic Publishers, Dordrecht, 1991.
[7]胡可云,陆玉昌,石纯一,粗糙集理论及其应用进展[J].清华大学学报(自然科学版),2001,41(1):64—68.
[8]史忠植.知识发现[M].清华大学出版社,中国计算机学会学术著作丛书230-264.143-168
[9]黄大荣,李劲.基于粗糙集理论的数据清洗模型[J].自动化技术与应用,2004,23(3):l1—17.
[10]苗夺谦.Rough Set理论中连续属性的离散化方法[J].自动化学报,2001,27(3):296-302.
[11] Jelonek Jacek, Krawiec Krzysztof, Slowinski Roman. Rough Set Reduction of Attributes and Their Domains for Neural Networks[J] . Computational Intelligence, 1995 11 (2): 339 - 347.
[12] Degang Chen,Qinghua Hu,etc.A new approach to attribute reduction of consistent and inconsistent covering decision systems with covering rough sets[J].Information Sciences,2007,177 (17):3500–3518.
[13] Ahn B S, Cho S S , Kim C Y. The Integrated Methodology of Rough Set Theory and Artificial Neural Network for Business Failure Prediction[J]. Expert Systems with Application, 2000, 18 : 65 - 74.
[14]陈淑珍.基于粗集的几种属性约简算法分析[J].武汉工业学院学报,2005,24(3):l18—120.
[15] Wang Jue,Miao Duoqian.Analysis on attribute reduction strategies of roughest[J].Journal of computer Science& Technology,1998,13(2):l89—193.
[16]李华,吴志强等.基于属性重要性的Rough集属性约简方法[J].计算机与现代化,2006,(12):69-70,74.
[17]吴明芬,许勇等.一种基于属性重要性的启发式约简算法[J].小型微型计算机系统,2007,28(8):1452-1455.
[18]朱颢东,钟勇.一种新的基于多启发式的特征选择算法[J].计算机应用,2009,29(3):849-851.
[19]田卫东,周创德等.基于简化分辨矩阵的粗糙集属性约简算法[J].计算机科学,2008,35(3):209-212.
[20]官礼和.基于可辨识矩阵的属性约简算法[J].计算机工程,2008,34(3):3-5.
[21]汪小燕,杨思春.一种基于分辨矩阵的新的属性约简算法[J].计算机技术与发展,2008,18(2):77-79.
[22]白燕娥,崔广才.基于遗传算法的属性约简算法研究与实现[J].长春理工大学学报,2005,28(3):36-38
[23]周明,孙树栋.遗传算法原理及应用.北京:国防工业出版社,2001.
[24]李敏强.遗传算法的基本理论与应用[M].北京:科学出版社, 2002.
[25] Zhou Yancong,Gu Junhua.A New Approach of Rough Attributes Discretization Based onGA[J].International Symposium on Test and Measurement,2005,(6):1-4.
[26]任永功,王杨等.基于遗传算法的粗糙集属性约简算法[J].小型微型计算机系统,2006,27(5):862-865.
[27]朱克敌,陶志.并行遗传算法在粗糙集属性约简中的应用[J].沈阳工程学院学报:自然科学版,2005,1(1):70-73.
[28]袁晓峰,许化龙等.基于量子遗传算法的粗糙集属性约简新方法[J].计算机工程,2007,33(15):184-186.
[29] K.-H.Han and J.-H.Kim, Genetic Quantum Algorithm and its Application to Combinatorial Optimization Problem[C]. in Proceedings of the 2000 Congress on Evolutionary Computation, pp. 1354-1360, July, 2000.
[30]李承祖.量子通信和量子计算[M].长沙:国防科技大学出版社, 2004.
[31] Narayanan A. Quantum computing for beginners. Proceedings of the 1999 Congress on Evolutionary Computation. Piscataway, NJ:IEEE Press, 1999, 2231-2238
[32]赵荣泳,张浩等.粗糙集连续属性离散化模型研究与应用要点分析[J].计算机工程与应用,2005,41(8):40-42,91.
[33]人工神经网络技术及应用王宏元史国栋主编中国石化出版社1-9页
[34]人工神经网络原理及仿真实例第2版高隽编著
[35] Simon Haykin.叶世伟译.神经网络原理[M].北京:机械工业出版社,2004.
[36]侯媛彬等编著.神经网络[M].西安电子科技大学出版社,2007.09.
[37]凌毅.神经网络在证券系统中的应用.北京工业大学硕士学位论文,2002.
[38]焦李成.神经网络系统理论.西安:西安电子科技大学出版社,1999:129-231.
[39] Haykin.S,叶世伟,史忠植译。神经网络原理[M].北京:机械工业出版社,2004(1):183—209.
[40]姚培福.人工神经网络在股票预测中的应用与研究[D].昆明理工大学,2007.03.
[41]林杰等,用神经网络方法预测股票短期走势,西南交通大学报[J],1998,(6):299-304
[42]王莎.BP神经网络在股票预测中的应用研究[D].中南大学,2008.04.
[43]智会强,牛坤,田亮,杨增军。BP网络和RBF网络在函数逼近领域内的比较研究[J].科技通报,2005,21(2):194-197.
[44] Yuhua Hou,Jian Cheng.Forecasting Coalmine Gas Concentration Based on RBF Neural Network[J].Information Acquisition,2007 International Conference:192-194.
[45] Dutta,G&Neeraj Mohan.Artificial Neural Network Models for Forecasting Stock Price Index in the Bombay Stock Exchanges[J].Journal of Emerging Market Finance,2006,5(3):283—295.
[46]王上飞,沈谦.径向基神经网络在股票预测中的应用[D].中国科学技术大学电子技术部,1998.
[47]周佩玲.神经网络结构设计的理论与方法.北京:国防工业出版社, 2003:163-221
[48]周佩玲.基于遗传算法的RBF网络用于股票短期预测.数据采集与处理,2003,1(1): 16-25
[49]朱明星,张德龙.RBF网络基函数中心选取算法的研究[J].安徽大学学报,2000,24(1)72—78.
[50]葛超,孙丽英,张淑卿,朱艺.RBF神经网络中心选取算法[J].河北理工大学学报,2007,29(4):95—97.
[51] Z.Ali,etc.Radial basis neural network for the classification of fresh edible oils using an electronic nose[J].Journal of Thermal Analysis and Calorimetry,2003,71(1):147-154.
[52]拉森,Microsoft SQL Server 2005 Reporting Services北京:清华大学出版社,2008.
[53]柴杰,江青茵.曹志凯.RBF神经网络的函数逼近能力及其算法[J].模式识别与人能,2002,15:310—316.
[54]徐晋.前馈神经网络学习新算法及其仿真.哈尔滨商业大学学报:自然科学版2004(1):1～3.
[55]付成宏,傅明[16]孙延风、梁艳春、孟庆福改进的神经网络最近邻聚类学习算法及其应用吉林大学学报2002 3 :63-66
[56]徐翔、黄道、李昱瑾一种改进RBF神经网络在股市建模及预测中的应用微型电脑应用2004 5(20):37-39
[57] Bao Rang Chang.Forecasting short—term stock price indexes—an integrated predictor vs.neural network predictor.TENCON’02.Proceedings.2002 IEEE Region 10 Conference on Computers,Communications,Control and Power Engineering,2002:817-820 v01.2
[58] Hiemstra,A stock market forecasting support system based on fussy logic.System Sciences,1994,3(1):281-287.
[59]董长虹编著。Matlab神经网络与应用[M].北京:国防工业出版社, 2005:15—60.
[60]刘维.精通MATLAB与c/c++混合程序设计[M].北京:北京航空航天大学出版社, 2005,6
[61]飞思科技产品研发中心. Matlab6.5辅助神经网络分析与设计.电子工业出版社,2003
[62]飞思科技产品研发中心.神经网络理论与Matlab7实现.北京:电子工业出版社,2005
[63]陈颓分形几何学地震出版社2005,2
[64]王波神经网络与时间序列模型在股票预测中的比较武汉理工大学学报(信息与管理工程版)2005、06
[65]陈遵德.Rough Set神经网络智能系统及其应用[J].模式识别与人工智能,1999,12(1):1-5
[66] Ruey S. Tsay.Analysis of Financial Time Series[M].潘家柱译.机械工业出版社,2005.
[67]郑丕谔,马艳华,基于RBF神经网络的股市建模与预测.天津大学学报, 2000 33(4):483—486
[68]朱赟,王行愚.RBF神经网络在股市趋势预测中的应用[J].华工理工大学学报,2002 28(5):547.
[69]叶中行,顾立庭。股市变化模式分类的两种神经网络方法[J].上海交通大学学报, 1995, 29 (2) : 10021041
[70]叶东毅,刘文标.径向基函数神经网络在股票走势模式分类中的应用[J].运筹与管理,1999,8(3):46-55.
[71]赵燕.基于神经网络的股票预测分析和研究[D].长安大学,2006.05
[72] Lin Jin-Cheng,Wu Kuo-Chiang.Design of Embedded Software Based on Rough Set and Neural Network[J].2007 International Conference on FSKD:141-145.
[73] Ting Chen,Jingqing Luo.Research on Rough Set-Neural Network and Its Application in Radar Signal Recognition [J].2007 International Conference,(8):3.
[74] Tian Ku.Design of a Novel Neural Networks Based On Rough Sets[J].Chinese Control Conference,2006,5:2078-2082.
[75] Cheung,W.S.,Ng,H.S.,Lam,K.P..Intraday stock price analysis and prediction.Management of Innovation and Technology,2000.ICMIT 2000-Proceedings of the 2000 IEEE International Conference,Singapore,2000:47—52, v01.1.
[76] Cao Zhiguo,Xiao Yang.The Application of Run-Length Features in Remote Sensing Classification Combined with Neural Network and Rough Set [J].2007 IEEE InternationalConference on Granular Computing:552.
[77] Yiwen Yang,Chaojnn Yang.Short term forecasting of stock market based on R/Sanalysi s and fuzzy neural networks.Systems,Man and Cybernetics,2003.IEEE International Conference,2003:2827—2832.
[78] Leigh,William,Purvis,Russell,Ragusa ,James M..Forecasting the NYSE composite index with technical analysis,pattern recognizer,neural network,and genetic algorithm:a ease study in romantic decision support.Decision Support Systems,2002 32(4):361-377.
[79] Qing Cao,Karyl B Leggio,Marc J Schniederjans. A Comparison between Fama and French's Model and Artificial Neural Networks in Predicting the Chinese Stock Market. Computes and Operations Research,2005(32):2499-2512.
[80] Ting Chen,Jingqing Luo.Research on Rough Set-Neural Network and Its Application in Radar Signal Recognition [J].2007 International Conference,(8):3
[81] Tian Ku.Design of a Novel Neural Networks Based On Rough Sets[J].Chinese Control Conference,2006,5:2078-2082.
[82] Deyou Xu.A Fault Diagnosis Method Combined Neural Network with Rough Set [J].International Conference on Signal Processin,2006,4(8):1829-1831.
[83]王天娥,叶徳谦,季春兰.粗糙集属性约简方法在股票预测中的应用研究[J].计算机工程与应用,2009,45(30).
[84]王京宝,径向基函数网络(RBF)在股市预测中的应用[J].科技信息,2007,02.
[85]马芳芳,王京宝.基于RBF神经网络的股票三类拐点的分析[J].科技信息(学术版)2008(6)
[86]王京宝.一种基于粗糙集与神经网络的股票决策支持系统的研究[辽宁工业大学硕士学位论文],2008:32-56.
[87]王天娥.基于粗糙集和RBF网络的股票时间序列分析研究[青岛理工大学硕士学位论文],2009:10-59