基于粗糙集與神經(jīng)網(wǎng)絡(luò)的股價(jià)走勢(shì)分析模型的研究.pdf

1、青島理工大學(xué)碩士學(xué)位論文基于粗糙集與神經(jīng)網(wǎng)絡(luò)的股價(jià)走勢(shì)分析模型的研究姓名：南敏申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用指導(dǎo)教師：葉德謙201012青島理工大學(xué)工學(xué)碩士學(xué)位論文AbstractSofarpeoplehavebeenlookingforallkindsofeffectivemethodstoavoidtheriskofstockmarketandtoobtainhigherreturnsfromstocks，somanytechn

2、ologiesforpretendingthetrendofstockmarkethavebeenproducedForthecomplexityofthestockmarketmanytechnologieshaveexposedtheirshortcomingsandinsufficiencyInthispapertheauthorbasesonthepreviousstudiesandusesroughsetsandneuraln

3、etworktopredictthetrendofstockpriceThemaincontendsareasfollows：(1)AtfirstobtaintheoriginaldatafromstockmarketusingDaZhihuiSoftware，andthenuseroughsetstoextractrepresentativedatafromoriginaldataInthisprocess，anewalgorithm

4、basedonquantumcomputingandgeneticalgorithmhasbeenusedandhassomeadvantagescomparedwithotheralgorithms(2)ConsideringthevariousadvantagesofRBFnetworktheauthorchoosesitforestablishingthepretendingmodel(3)Astherealemanylearni

5、ngmethodsforRBFnetworkSOchoosingtheeffectivemethodhasbeenthefocusInthisPapertheauthorcomparesfouralgorithmsandchoosesthebestfromthemthroughexperiments(4)Atlast，theauthorestablishesamodelbasedonthefrontresearchTheinputvec

6、torsarechoosedfromtheroughsetsresults；theoutputvectorisdividedsixkindanddrawnedintheresultcharttohelpUSerSmakedecisionsmoreeffectivelyFinallytheauthorUSgSrepresentativedatatovalidatethecorrectnessandcredibilityofthemodel