English  |  正體中文  |  简体中文  |  Items with full text/Total items : 70585/70585 (100%)
Visitors : 23104725      Online Users : 700
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version


    Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/78735


    Title: 階層式分類與回歸法用於數值預測;Soft Estimation by Hierarchical Classification and Regression
    Authors: 柯士文
    Contributors: 國立中央大學資訊管理學系
    Keywords: 數值預測;資料探勘;分類;回歸;階層式預估;Prediction;data mining;classification;regression;hierarchical estimation
    Date: 2018-12-19
    Issue Date: 2018-12-20 13:46:21 (UTC+8)
    Publisher: 科技部
    Abstract: 本研究是最近在“Neurocomputing”期刊上發表的一項研究之延伸,計劃主持人Shih-Wen Ke柯士文為第一作者(Ke,et al, 2017, Soft estimation by hierarchical classification and regression. In Neurocomputing, Vol. 234, pages 27-37),其中提出了分層分類和回歸(HCR)方案。過去提出的分層估計方法,通常是多個估計模型的組合,用於解決一些特定的領域問題。然而,在文獻中,沒有通用的分層方法進行估計,也沒有將分類和估計技術分層結合的基於混合的方法。因此,HCR是我們最近提出的方法。 HCR方法在平均絕對百分比誤差(MAPE)和均方根誤差(RMSE)率方面明顯優於三個常用的單一層預測模型,即線性回歸,多層感知器神經網絡和支持向量回歸。另外,發現使用基於基因演算法GA的數據預處理方法將訓練集分類為4個子集是最佳值(即,k = 4),並且4類SVM + MLP優於3個基線分層回歸模型。因此我們看到做為一個通用的分層回歸方法(HCR)的潛力,我們會繼續投入時間,改善HCR的預測表現,並在選定的範疇和應用範圍內嘗試HCR方法。我們提出的計劃和一些可以預見的研究問題總結如下:在處理包含大量數據樣本和/或屬性的大規模回歸問題時,將結合特徵選擇方法到HCR中。我們最近的發現顯示,MLP神經網絡是線性回歸和支持向量回歸中表現最好的回歸模型。這個觀察導致我們提出一個新的問題:“其他的神經網絡架構能否用於HCR?”因此,我們將研究如何將深度神經網絡架構合併到HCR中。 ;This study is an extension to a recent work published in the journal of Neurocomputing, first-authored by the Principal Investigator Shih-Wen Ke 柯士文 (Ke, et al., 2017, Soft estimation by hierarchical classification and regression, In Neurocomputing, Vol. 234, Pages 27-37) where a hierarchical classification and regression (HCR) scheme was proposed. Hierarchical estimation approaches, usually a combination of multiple estimation models, have been proposed for solving some speci?c domain problems. However, in the literature, there is no generic hierarchical approach for estimation and no hybrid based solution that combines classi?cation and estimation techniques hierarchically. Hence the HCR was proposed in our recent work. the HCR approach significantly outperformed three well-received single flat prediction models, i.e. linear regression, multilayer perceptron neural network and support vector regression in terms of mean absolute percentage error (MAPE) and root mean squared error (RMSE) rates. In addition, it is found that using the GA-based data pre-processing approach to classify the training set into 4 subsets is the best threshold (i.e., k=4) and the 4-class SVM+MLP outperforms three baseline hierarchical regression models.Having seen the potential of the proposed HCR as a generic hierarchical regression scheme, we will continue to invest time in improving effectiveness of the HCR and to employ the HCR in selected domains and applications. Our proposed plans and some foreseeable research questions are summarized as below. Feature selection methods will be incorporated when dealing with larger scale regression problems that contain very large numbers of data samples and/or attributes. Our recent finding showed that the MLP neural networks was the best performing regression model amongst linear regression and support vector regression. This observation leads us to a new question that “Can other neural network architectures be employed in the HCR?” Therefore, we will study how to incorporate deep neural networks architectures into the HCR.
    Relation: 財團法人國家實驗研究院科技政策研究與資訊中心
    Appears in Collections:[資訊管理學系] 研究計畫

    Files in This Item:

    File Description SizeFormat
    index.html0KbHTML142View/Open


    All items in NCUIR are protected by copyright, with all rights reserved.

    社群 sharing

    ::: Copyright National Central University. | 國立中央大學圖書館版權所有 | 收藏本站 | 設為首頁 | 最佳瀏覽畫面: 1024*768 | 建站日期:8-24-2009 :::
    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback  - 隱私權政策聲明