To search, Click
below search items.
|
|
![](../image/sub_title/01_title_01.gif)
All
Published Papers Search Service
|
Title
|
Exploring and Determining Missing-data Imputation Method for Socio-economic Data by Way of Designing a Simulation Study in the Context of Gross National Happiness (GNH) Data Set
|
Author
|
Sonam Tshering, Takeo Okazaki, Satoshi Endo
|
Citation |
Vol. 12 No. 5 pp. 32-38
|
Abstract
|
This paper proposes a bootstrap-based EM method of imputing missing data in socio-economic data set. The method is proposed by way of designing a simulation study in the context of GNH data set. The data set is explored for significant latent factor(s) and manifest variables. A model is fitted to it and based on the model, pseudorandom data with specific missingness mechanism are generated. The generated missing data are imputed with both traditional and proposed methods under MCAR, MAR and MNAR missingness mechanism. The model estimates and the MASE values of the proposed method are compared with that of the traditional methods to assess the accuracy of the proposed method. The method is then applied to incomplete GNH data set after ascertaining its validity.
|
Keywords
|
Bootstrap, Missingness, Imputation, Algorithm
|
URL
|
http://paper.ijcsns.org/07_book/201205/20120506.pdf
|
|