簡易檢索 / 詳目顯示

回結果列表
研究生： 馬彬立
Billy Malherbe
論文名稱： 將雙重機器學習應用於結構方程模型
Applying Bootstrap and Double Machine Learning to Structural Equation Modeling
指導教授： 雷松亞
Ray, Soumya
口試委員: Yoo, Jaewon
Yoo, Jaewon
Danks, Nicholas
Danks, Nicholas
學位類別： 碩士
Master
系所名稱： 科技管理學院 - 國際專業管理碩士班
International Master of Business Administration(IMBA)
論文出版年： 2022
畢業學年度： 110
語文別： 英文
論文頁數： 32
中文關鍵詞： 雙重機器學習 、自舉 、結構方程模型
外文關鍵詞： Double Machine Learning, SEM-PLS, Bootstrapping
相關次數： 點閱：3  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • N/A

    Partial Least Square - Structural Equation Modeling (PLS-SEM) has become a quasi-standard technique in management research and other fields. We are investigating techniques found in regression that could be replied to PLS-SEM. We are taking the result of the non-parametric bootstraps to compare with the other type of bootstraps, residual and wild bootstraps. We also call the resulting set of methods double or debiased ML (DML), which is a method that can solve the problem of regularization bias and overfitting. We will explore what these techniques are and implement them for PLS-SEM

    ACKNOWLEDGMENT 1 INDEX OF FIGURES 6 I. INTRODUCTION 1 II. LITERATURE REVIEW 3 1. PLS-SEM(PARTIAL LEAST SQUARES - STRUCTURAL EQUATION MODELING) 3 1.1. What is PLS-SEM? 3 1.2. Where PLS-SEM is used? 4 Non-normal data 4 Small sample size 4 Formative indicators 4 2. BOOSTRAPPING 5 2.1. DIFFERENT TYPES OF BOOTSTRAPS AND TECHNIQUES OF BOOTSTRAPPING 6 2.1.1. General types of bootstrapping 6 Nonparametric bootstrap 6 Semiparametric bootstrap 7 Parametric bootstrap 7 Why not parametric and semiparametric bootstrap for PLS? 8 2.1.2. Alternative Techniques of Non-parametric bootstrapping 8 - Residual Bootstrap 8 Wild Bootstrap 9 2.2. BOOTSTRAPPING IN PLS-SEM 11 2.2.1. Nonparametric bootstrap in PLS-SEM 11 Number of Bootstrap samples 11 2.2.2. Alternative Bootstraps in PLS-SEM (Residual and Wild bootstrap) 12 3. DOUBLE MACHINE LEARNING (DML) 12 3.1. Unbiasing Regression Using DML 12 3.1.1. Where DML is used? 13 3.1.2. Machine Learning or Double Machine Learning: 14 3.2. Complex PLS Models 14 3.2.1. Multiple dependent variables, with mediation and moderation 14 3.2.2. New ways of improving causal inference using Double Machine Learning (DML) 15 III. METHODOLOGY 16 1. APPLYING NEW METHODS TO PLS-PM 16 2. ALTERNATIVE BOOTSTRAP METHODS FOR PLS 16 2.1 DATA 16 2.2 ORDINARY NON-PARAMETRIC BOOTSTRAP 18 2.3 RESIDUAL BOOTSTRAP 20 2.3.1 ADAPTED ALGORITHM (PSEUDOCODE) 20 2.3.2 BOOTSTRAP RESULTS 21 2.4 WILD BOOTSTRAP 22 2.4.1 ADAPTED ALGORITHM (PSEUDOCODE) 22 2.4.2 BOOTSTRAP RESULTS 23 3. DOUBLE MACHINE LEARNING FOR PLS 24 3.1. CROSS-FITTED-DML FOR PLS 24 3.1.1. ADAPTED ALGORITHM (PSEUDOCODE) 24 3.1.2. DML RESULTS 25 3.2. MONTE-CARLO-CROSS-FITTED FOR PLS 26 3.2.1. MONTE CARLO SIMULATIONS 26 3.2.2. ADAPTED ALGORITHM (PSEUDOCODE) 27 IV. DISCUSSION AND CONCLUSION 29 REFERENCES 31

    o Chernozhukov, V., Chetverikov, D., Demirer, M., Duflo, E., Hansen, C., Newey, W., & Robins, J. (2018). Double/debiased machine learning for treatment and structural parameters.
    o Shah, R., Goldstein, S. M., Unger, B. T., & Henry, T. D. (2008). Explaining anomalous high performance in a health care supply chain. Decision Sciences, 39(4), 759-789.
    o Beebe, D. W., and Gozal, D. (2002). Obstructive sleep apnea and the prefrontal cortex: towards a comprehensive model linking nocturnal upper airway obstruction to daytime cognitive and behavioral deficits. J. Sleep
    o Lohmöller,J.B. and Wold, H. (1980), "Three-mode path models with latent variables and partial least squares (PLS) parameter estimation", paper presented at European Meeting of the Psychometric Society, Groningen, The Netherlands.
    o Ray, S., Ow, T., & Kim, S. S. (2011). Security assurance: How online service providers can influence security control perceptions and gain trust. Decision Sciences, 42(2), 391-412.
    o Hair, J. F., Hult, J. G. T. M., Ringle, C. M., & Sarstedt, M. (2014). A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM).
    o Beran, R. (1986). Discussion of Wu, CFJ: Jackknife, bootstrap, and other resampling methods in regression analysis (with discussion). Ann. Statist, 14, 1295-1298.
    o K. G. Jöreskog, H. O. A. W. (1982). The ML and PLS techniques for modeling with latent variables: historical and comparative aspects. In Systems Under Indirect Observation (pp. 263–270).
    o Davison, A., and D. Hinkley. 1997. Bootstrap Methods and Their Application. Cambridge: Cambridge University Press.
    o Efron, B. (1982). The jackknife, the bootstrap, and other resampling plans..
    o Barclay, L. M., & Lloyd, B. (1996). The misery of motherhood: alternative approaches to maternal distress. Midwifery, 12(3), 136-139.
    o Ray, S., Danks, N., & Calero Valdez, A. (2021). SEMinR: Domain-specific language for building, estimating, and visualizing structural equation models in R. Estimating, and Visualizing Structural Equation Models in R (August 6, 2021)
    o Jeong, J., & Maddala, G. S. (1993). 21 A perspective on application of bootstrap methods in econometrics.
    o MacKinnon, J. G. (2006). Bootstrap methods in econometrics. Economic Record, 82, S2-S18.
    o Sharma, J. L., Mougoue, M., & Kamath, R. (1996). Heteroscedasticity in stock market indicator return data: volume versus GARCH effects. Applied Financial Economics, 6(4), 337-342.
    o Skoufias, E., & Vinha, K. (2021). Child stature, maternal education, and early childhood development in Nigeria. PloS one, 16(12), e0260937.

    QR CODE