首页 > 资料专栏 > 论文 > 财税论文 > 金融投资论文 > 基于投资者画像的股票个性化推荐模型研究_MBA毕业论文PDF

基于投资者画像的股票个性化推荐模型研究_MBA毕业论文PDF

八维空间
V 实名认证
内容提供者
资料大小:11114KB(压缩后)
文档格式:PDF(182页)
资料语言:中文版/英文版/日文版
解压密码:m448
更新时间:2024/2/2(发布于广东)
阅读:9
类型:金牌资料
积分:--
推荐:升级会员

   点此下载 ==>> 点击下载文档


“基于投资者画像的股票个性化推荐模型研究_MBA毕业论文PDF”第1页图片 “基于投资者画像的股票个性化推荐模型研究_MBA毕业论文PDF”第2页图片 “基于投资者画像的股票个性化推荐模型研究_MBA毕业论文PDF”第3页图片 图片预览结束,如需查阅完整内容,请下载文档!
文本描述
# 以下为纯文本简介,不带文件格式;详细内容请下载文档查看;
M 论文题目:基于投资者画像的股票个性化推荐模型研究 学科专业:工商管理 :名: 研究生段刚龙签 :: 指导教师扈文秀教授签名 : 杨熙安教授签名 摘要 、 伴随着我国金融市场的迅猛发展,人工智能机器学习和大数据技术的极大进步,许 多金融市场的经营和盈利腏椒⑸朔旄驳氐谋浠蹲收哐扒舐闫渥什V翟鲋档? 理财求也不断攀升,实现差 需异化股票投资交易风险和收益偏好的个性化需求日趋明显。 智能投资顾问的不断发展为满足投资者的个性化金融需求提供了机会。作为资产管理前沿 领域的智能投资顾问,智能投资顾问可以智能分析投资人的风险偏好,并通过算法膒突? 人善, 得投资最完的资产配置为投资者购买股票提供了个性化的选择空间,而提高智能投 习 顾学效率和准确性的个性化推荐技术则是智能投顾中的重中之重。协同过滤推荐已经成 的, 为了智能投顾首选算法但存在着数据稀疏性、冷启动、长尾效应导致股票难以被推荐 , 问,, 等题导致推荐的性能不高用户的体验较差因此,破解智能投顾中股票的个性化推 荐方法与技术的难题,实现股票投资者的精准化的理财服务具有重要的理论价值与现实意 义。 “ , 针对以上问题论文结合用户画像技术,围绕股票投资者个性化需求的关键推荐算 ” 法的基本问题展开研宄,首先,针对股票投资者的投资目的和过程进行解析,构建了投 资者建腄?椤⒐善倍韵蠼D?椤⑼萍鏊惴?榈难芯克悸汾此基础上,构建股票投 资者智能投顾的用户画像,设计了用户画像的事实标签、分类膒捅昵┖推兰勰p捅昵┨? 系,采用XGBoost算法构建了投资者的用户分类膒停⒉扇OpSIS法对股票投资者 标签进行了评价浯危ü诠亓嬖颉⒒谖谋灸谌莺蜕疃刃耸咏牵菇? 3种情景下的个性化推荐子模型,采用关联规则实现股票行业推荐,在股票行业推荐的基 础上实现了个股推荐,从S于股票评论及金融妍件的文本内界视角.构迚文本数据的金融 事件词典,提出蕋 于结构化信息股票盈利预估校則和多任务股票盈利预估模型,进而进行 股票盈利计算及结合用户画像筛选符合⑴户偏好的股栗.设计了数据预处理层、子 - 推荐算法层、推荐算法融合层和膒托砥兰鄄愕幕旌贤萍隹蚣埽贚ZApriori、MSEEM 和FCM子模型分析的站础上,构建了混合多专家网络的股票推狞融合算法,并采用算例 实验对膒退惴ǖ能扌越辛搜橹ぃ? 本文将从以下三个方而介绍创新点: I 西安理工大学博士学位论文 首先,基于投资者行为偏好视角,构建了股票投资者的用户画像标签体系设计和膒汀? 现有研究虽然识别了大量的股票投资者购买偏好,但是依旧虯粼诖车耐蹲首楹辖ㄒ樯希? 而用户画像的相关研究则聚焦于互联网用户的网络行为。该研究将股票行业的用户投资偏 好和用户画像技术跨界融合,采用股票的智能投顾的逻辑架构,建立包括数据采集、数据 挖掘及过滤和标签提取及重组的用户画像过程,从投资能力标签、行为特征标签、行业偏 好标签、地域偏好标签与风险偏好标签5个方面构建了股票投资用户的画像标签体系,涵 盖了事实标签、分类膒捅昵┖推兰勰p捅昵┶此基础上,基于Gradient Boosting、 TOPSIS和FND-LDA2vec 的算法优势,构建了投资者用户分类标签膒汀⑵兰郾昵┠p? 和股吧话题偏好挖掘膒停⑼ü憷治鲅橹ち四p偷挠行浴? 其次,基于关联规则、基于股票内容和深度协同过滤视角,分别构建了投资者股票智 能推荐子膒汀D壳把芯咳嗽币丫范斯亓嬖颉⒒谀谌莺托说耐萍龇绞胶图? 术的优点,但是算法仍然具有数据稀疏性、冷启动、长尾效应等问题。本研究将三种算法 分别进行了改进,针对关联规则利用其信息的联动优势,挖掘股票行业和股票指数的内部 关联的行业联动与个股涨跌趋势,同时兼顾两个层面的信息实现基于改进Apriori的股票 推荐;针对基于内容推荐股票评论及金融事件,提出基于结构化信息股票盈利预估膒秃? 多任务股票盈利预估膒停扑愎善庇耐笔迪钟没闷ヅ洌徽攵孕说氖? 稀疏性问题,通过股票池的模糊聚类和多阶段匹配,结合深度学习算法优化近邻协同过滤 算法生成个性化股票推荐列表。 最后,构建了数据预处理层、子推荐算法层、推荐算法融合层和膒托Ч兰鄄愕娜? 合推荐体系,以及混合多专家网络的股票推荐融合算法钟械囊丫佣嘟嵌冉⒘烁鲂? 化的混合选择方法,不过没有运用到智能投顾的股票选择领域,并且选择方法也是基于数 据的混合,忽略了过程与体系的兼容性。本方法立足于股票市场中智能投顾的股票配置的 研究和服务的总体要求,基于关联规则的LZ-Apriori算法、基于内容的MSEEM算法和 基于深度协同过滤的FCM算法,搭建起基于混合多专家网络的股票推荐融合算法体系, 构建了生产合成数据、混合多专家系统、以及股票推荐融合(输入嵌入、多专家编码、门 控网络和交互输出)的多阶段处理流程,并通过算例实验验证了算法的有效性。 关键词:智能投顾;股票选择与推荐;协同过滤;投资偏好;用户画像 研究类型:应用基础研究 II t Ab st r u c t c o m m e n d a io nMe th o d Tit l e :Re s e a rc h o n p e rs o n a l iz e dSt o c kRe Bas e d o n I n v e s t o rp o rtr a it r :Bu s in ira io n Ma o es s Ad m n is t t j u r e am e :Ga n l o n u a Sin a t : N g gDn g io r in a u re : Su e rv s :p r of.We n x iu u St p Hg ’ r p ro f .Sin a tu e : Xia nYa n g g Ab s t ra c t e ' Wih the r a id d ve o e nd h e re r s t l t il r t t t r rt i m n o f h n a s na nc ia m a ke a a o e s o f a icia p p Cfig p g f l it l l iil i n e en ce m ach n e e arn n an d b id h e o e r at in r f it del f g ,g at a t e chn o l o t o n a d o m o o m an g gy ,p p y ' tra it ia l ie it e d o n bus n e ss s n h fia n c ia l er n e v e chan e s n m arke th av eun d go sub ers iv g .In ve st o rs fin anc ial needsto eettheir asset re erv ation and a reciation are also risin an m p s pp g ,d the er n alied de and o p so z mt realize differentiated stock invest ent tradin risk an d return mg referen ceisb e in i p comg ncreasin l obviou s.The continuousrise o fitellientinvest en g y t ng m ' adviser h a id s s rov edo ortu nitiesforinvestors ersonalizedfinancialneeds.Asanin ellient p p p p t g ies e d nv tmnta visor serviceintheasset ana e entindu str intellientinvest en advise y mg m y ,g mt rs ' can in tellientl evaluate investors risk referen ces an d obtain inv st g y p e mentcombin ation su estio atchin inv estors based on l t gg n s mg a orih odels rovidin in vestors ith g mm,p g w ersonalizedch oices acefor urchasin stocks personalized eco da i p p p g ,r mmen t o ntechn olo to gy ir v mp o ethe learnin efficienc andaccurac ofin tellient invest entadvisers isthe to g y y g m p riorit ofite itin es entad isers s he re e t p y n ll gen v tmv .At p f rredal orih forintellientinv est ent g m g m adv isers collab orativefilterin reco en dationfaces roble ssuchasdatas ar it colds art ,g mm p m p s y ,t , an dd iffiult irec endin lon a te s resultin n o reco en a c y nommg g t ilim,g il wmmd tion erfor a ce p mn an d oor serex erience h erefore iisof reattheoreticalvalueand racticalsin p u p .T,t g p g ificanceto crack h e erson alized reco en da on e hods an d technolo ies of cks n t t p mmtimt g sto iin tellien g inv estmen tad visersandrealizetheaccuratefinancialservicesofstockinvestors. I ie of he b bl h erco bines theu er r r iec n an d nv t a ove pro e s,t e pap m s p o t a tt h olo w m gy " fo c ses h b iroble of ke reco endational orith forthe ersonalids of u ont e as cp my mm g mp zedn ee " stockinv es ors .rs h e a ercon structsth eu ser ortraitofth eintellientinvest en adviser t Fit,t p p p g mt ofstockid sins threesta esoftheuser ortraitcon struction rocess constru ctsthe nv estors,e g g p p , in t t a e th rou h the co llection of ortrait indica ors and dex s ste ofthe user or rail b l g p t , y mp oost a or th is used to bu calcu lates the index eiht u sing AHp ,Xgb lg im ild the user wg th e ers ectiveofasso ciati classificationmodelo finv estors.Secondly ,fromp p onrules,stock ill 西安理工大学博士学位论文 content and deep collaborative filtering,personalized recommendation sub algorithms under three scenarios are constructed.Association rules are used to analyze the linkage effects between stocks and between industries,identify the related indicators of stock fluctuations,and extract the stock characteristics from the unstructured information of the sector,basic information and comment information based on stock content,Based on the deep collaborative filtering,the users preference intensity of different features is identified,and the multi-stage stock matching is carried out by combining the fuzzy clustering results.Finally,the stock hybrid recommendation algorithm is designed to realize the logical framework of data preprocessing layer,sub recommendation algorithm layer,recommendation algorithm fusion layer and model effect evaluation layer.The multi-stage fusion of the early,middle and late stages of the algorithm is adopted,and an example is used to verify the effectiveness of the model algorithm. The innovation of this paper is mainly reflected in the following three aspects: Firstly,based on the perspective of investor behavior preference,a user portrait model of intelligent investment adviser for stock investors is constructed.Although the existing studies have identified a large number of stock investorspurchase preferences,they still stay in the traditional portfolio recommendations,while the research on user portraits focuses on the network behavior of Internet users.This research integrates the user investment preference in the stock field with the user portrait technology.Based on the logical framework of the stock based intelligent investment adviser,the user portrait process of data collection,data mining, filtering,label extraction and reorganization is constructed.The portrait label system of stock investment users is constructed from five aspects:investment ability label,behavior feature label,industry preference label,regional preference label and risk preference label.On this basis,based on the algorithm advantages of Gradient Boosting,TOPSISand FND-LDA2vec, this paper constructs the investor user classification tag model,evaluation tag model and stock bar topic preference mining model,and verifies the effectiveness of the model through numerical example analysis. Secondly,based on association rules,stock content and in-depth collaborative filtering,the sub algorithms of investor stock intelligent recommendation are constructed respectively. Although the existing research has identified the advantages of association rules,content-based and collaborative filtering recommendation methods and technologies,the algorithm has some problems,such as data sparsity,cold start,long tail items and so on.In this study,the three algorithms are improved respectively.Aiming at the association rules,we use their information linkage advantages to mine the industry linkage and individual stock fluctuation trend of the internal association of the stock industry and the stock index,and take into account the two IV 。。。以下略