English  |  正體中文  |  简体中文  |  全文笔数/总笔数 : 62567/95223 (66%)
造访人次 : 2523949      在线人数 : 32
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
搜寻范围 查询小技巧:
  • 您可在西文检索词汇前后加上"双引号",以获取较精准的检索结果
  • 若欲以作者姓名搜寻,建议至进阶搜寻限定作者字段,可获得较完整数据
  • 进阶搜寻

    jsp.display-item.identifier=請使用永久網址來引用或連結此文件: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/105321

    题名: 巨量資料探討及分析 : 以台灣50 ETF為例
    其它题名: A study and research on big data analysis- an example to Taiwan 50 ETF
    作者: 陳昱安;Chen, Yu-An
    贡献者: 淡江大學管理科學學系碩士班
    关键词: 大數據;台灣50 ETF;技術指標;倒傳遞類神經網路;Big Data;Taiwan 50 ETF;Technical Indicators;Back-propagationNeural Network
    日期: 2015
    上传时间: 2016-01-22 14:53:11 (UTC+8)
    摘要: 「大數據」目前為實務界非常重視的議題之一。因電腦儲存量變大、處理速度快、網路寬頻化及行動網路的普及而產生許多的資料及數據,造就了大數據時代的來臨。政府及企業也受到大數據的衝擊,提出了許多政策及方案,政府希望藉由大數據的分析,提高政府的效能及效率並且增加人民的福祉。而企業希望藉由大數據使產業升級,提升企業績效。
    ETF為目前最受歡迎的投資工具之一。ETF因透明度高、投資成本低、投資組合多元化且買賣方便等優點,受到投資人的青睞。本研究以模擬大數據的研究觀點,找出與台灣50 ETF股價具相關性非結構化數據,透過非結構化數據,使投資人更了解台灣50 ETF之走勢。
    實證結果:透過牛頓法改良倒傳遞類神經網路(LM-BP)之MAPE=1.881.88%、RMSE=2.2091、R^2=0.889051,顯示本研究股價預測精確。運用多變量分析之迴歸模型找出與本研究台灣50 ETF股價預測相關性非結構化數據為:(1) 公益彩
    Big data" is an serious issue in data analysis. Due to the large amount of computer storage, processing speed, broadband network technology and the popularity of mobile networks, we face a large number of information which desire a new age of big data. Therefore, public office need to make new policies and programs and improve the effectiveness and efficiency of government and increases the welfare of the people by applying of Big Data, most of companies want to use big data for industry promotion and enhancing enterprise performance. Due to ETF has high transparency, low investment costs, portfolio diversification and facilitate trading, ETF is one of the most popular investment instruments. In this study, simulation of Big data view, we find Taiwan 50 ETF shares price a correlation analysis through unstructured data, thus enabling investors to better understand the trend of Taiwan 50 ETF. The analysis results show that the revised result back-propagation neural network has the smaller forecasting error with MAPE=1.881.88% and RMSE=2.2091, shows the study accurately predict stock price. Using multivariate analysis of regression models to identify with this study in Taiwan 50 ETF shares predictive correlation unstructured data: (1) lottery sales (2) people travel abroad (3) the number of domestic and imported wine (4) out-eater price of CPI.
    显示于类别:[管理科學學系暨研究所] 學位論文


    档案 描述 大小格式浏览次数



    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - 回馈