Spatial Estimation of Finite Population Total under Geographically  Weighted Regression using Forward Stepwise Variable Selection

Nobin Chandra Paul; Anil Rai; Tauqueer Ahmad; Ankur Biswas; Prachi Misra Sahoo

doi:10.56093/jisas.v78i03.171421

Authors

Nobin Chandra Paul ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Anil Rai Indian Council of Agricultural Research, New Delhi
Tauqueer Ahmad ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Ankur Biswas ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Prachi Misra Sahoo ICAR-Indian Agricultural Statistics Research Institute, New Delhi

https://doi.org/10.56093/jisas.v78i03.171421

Keywords:

Akaike information criterion; GWR; spatial estimator; spatial non-stationarity.

Abstract

Unlike ordinary least square model, the geographically weighted regression model takes into account spatial non-stationarity and can capture the spatially varying relationship between several variables. Although, a particular model should contain all pertinent covariates but too many insignificant covariates make the model unnecessarily complex. Therefore, it is important to choose important covariates having significantly high correlation with the study variable. Here, a forward stepwise variable selection procedure under the geographically weighted regression model framework has been proposed for choosing significant covariates and compared with the existing forward stepwise ordinary least square method. Further, an estimator of finite population total incorporating spatial information has been developed. The performance of the proposed spatial estimator was compared empirically under both forward stepwise geographically weighted regression and forward stepwise ordinary least square method through a spatial simulation study. It was found that the performance of the spatial estimator using forward stepwise geographically weighted regression method is
better than the forward stepwise ordinary least square method.

Downloads

Download data is not yet available.

References

Ahmad, T., Bhatia, V.K., Sud, U.C., Rai, A., and Sahoo, P.M. (2013). Study to develop an alternative methodology for estimation of cotton production, Project Report, IASRI publication.

Ahmad, T., Sud, U.C., Rai, A., and Sahoo, P.M. (2020). An alternative sampling methodology for estimation of cotton yield using double sampling approach. Journal of the Indian Society of Agricultural Statistics, 74(3), 217-226.

Anselin, L. (1995). Local Indicators of Spatial Association-LISA. Geographical Analysis, 27, 93-115.

Anselin, L. (1996). The Moran Scatter Plot as an ESDA Tool to Assess Local Instability in Spatial Association, In: M. Fischer, H. Scholten and D. Unwin, Eds., Spatial Analytical Perspectives on GIS in Environmental and Socio-Economic Sciences, 111-125.

Biswas, A., Rai, A., Ahmad, T. and Sahoo, P.M. (2017). Spatial estimation and rescaled spatial bootstrap approach for finite population. Communication in Statistics- Theory and Methods, 46, 373-388.

Biswas, A., Rai, A. and Ahmad, T. (2020). Spatial bootstrap variance estimation method for missing survey data. Journal of the Indian Society of Agricultural Statistics, 74(3), 227-236.

Brunsdon, C., Fotheringham, A.S. and Charlton, M.E. (1996). Geographically weighted regression: a method for exploring spatial non-stationarity. Geographical Analysis, 28, 281-298.

Brunsdon, C., Fotheringham, S. and Charlton, M. (1998). Geographically weighted regression-modelling spatial non-stationary. The Statistician, 47(3), 431-443.

Cressie, N.A.C. (1991). Statistics for spatial data. Wiley, New York.

Comber, A., Wang, Y., Lu, Y., Zhang, X. and Harris, P. (2018). Hyper local geographically weighted regression: extending GWR through local model selection and local bandwidth optimization. Journal of Spatial Information Science, 17, 63-84.

Comber, A. and Harris, P. (2018). Geographically weighted elastic net logistic regression. Journal of Geographical Systems, 20, 317-341.

Fotheringham, A.S., Charlton, M.E. and Brunsdon, C. (1998). Geographically weighted regression: a natural evolution of the expansion method for spatial data analysis. Environment and Planning A, 30, 1905-1927.

Fotheringham, A.S., Brunsdon, C., and Charlton, M. (2002). Geographically weighted regression: the analysis of spatially varying relationships. John Wiley & Sons Ltd, England.

Fotheringham, A.S., Yang, W. and Kang, W. (2017). Multiscale Geographically Weighted Regression (MGWR). Annals of the American Association of Geographers, 107, 1247-1265.

Gollini, I., Lu, B., Charlton, M., Brunsdon, C. and Harris, P. (2015). GWmodel: an R Package for exploring Spatial Heterogeneity using Geographically Weighted Models. Journal of Statistical Software, 63(17), 1-50.

Gujarati, D.N., Porter, D.C., and Gunasekar, S. (2012). Basic Econometrics. McGraw-Hill Education, 5th edition.

Leung, Y., Mei, C.L. and Zhang, W.X. (2000). Statistical tests for spatial non-stationarity based on the geographically weighted regression model. Environment and Planning A, 32(1), 9-32.

Lu, B., Charlton, M., Harris, P., and Fotheringham, A. (2014). Geographically weighted regression with a non-Euclidean distance metric: a case study using hedonic house price data. International Journal of Geographical Information Science, 28(4), 660-681.

Middya, A.I. and Roy, S. (2021). Geographically varying relationships of COVID-19 mortality with different factors in India. Scientific Reports, 11, 7890.

Moran, P.A.P. (1948). The Interpretation of Statistical Maps. Journal of the Royal Statistical Society, Series B (Methodological), 10(2), 243-251.

Nakaya, T., Fotheringham, A.S., Charlton, M., and Brunsdon, C. (2009). Semiparametric geographically weighted generalised linear modelling in GWR 4.0. In: 10th International Conference on GeoComputation, 30th November-2nd December 2009, UNSW, Sydney.

Paul, N.C., Rai, A., Ahmad, T., Biswas, A. and Sahoo, P.M. (2022). Bootstrap Variance Estimation of Spatially Integrated Estimator of Finite Population Total in Presence of Missing Observations. Journal of Community Mobilization and Sustainable Development, 17(3), 1039-1048.

Paul, N.C., Rai, A., Ahmad, T., Biswas, A., and Sahoo, P.M. (2023a). GWR-assisted integrated estimator of finite population total under two-phase sampling: a model-assisted approach. Journal of Applied Statistics, 51(12), 2326-2343. https://doi.org/10.1080/02664763.2023.2280879

.

Paul, N.C., Rai, A., Ahmad, T., Biswas, A., and Sahoo, P.M. (2023b). Spatial approach for the estimation of average yield of cotton using reduced number of crop cutting experiments. Current Science, 125(5), 518.

Paul, N.C., Rai, A., Ahmad, T., Biswas, A., and Sahoo, P.M. (2024). Spatially integrated estimator of finite population total by integrating data from two independent surveys using spatial information. Journal of the Korean Statistical Society, 53, 222-247. https://doi.org/10.1007/s42952-023-00244-1

.

Paul, N.C., Rai, A., Ahmad, T., and Biswas, A. (2024). Integration of Spatial Data from Two Independent Surveys: A Model-Based Approach Using Geographically Weighted Regression. Journal of the Indian Society for Probability and Statistics, 25, 895-921. https://doi.org/10.1007/s41096-024-00212-w

.

Pebesma, E.J. (2004). Multivariable geostatistics in S: the gstat package. Computers and Geosciences, 30(7), 683-691.

Royall, R.M. (1970). On finite population sampling theory under certain linear regression models. Biometrika, 57(2), 377-387.

Royall, R.M. (1978). An empirical study of prediction theory in finite population sampling: simple random sampling and the ratio estimator. Survey Sampling and Measurement. Academic Press.

Royall, R.M., and Cumberland, W.G. (1981). The finite-population linear regression estimator and estimators of its variance-An empirical study. Journal of the American Statistical Association, 76, 924-930.

Santiago, B. (2010). Generating spatially correlated random fields with R. Link: (http://santiago.begueria.es/2010/10/generating-spatially-correlated-random-fields-with-r/

)

Saha, B., Biswas, A., Ahmad, T., and Paul, N.C. (2023). Geographically Weighted Regression-Based Model Calibration Estimation of Finite Population Total Under Geo-referenced Complex Surveys. Journal of Agricultural, Biological and Environmental Statistics, 29, 793-811. https://doi.org/10.1007/s13253-023-00576-9

.

Saha, B., Biswas, A., Ahmad, T., Misra Sahoo, P., Aditya, K., and Paul, N.C. (2024). Geographically weighted regression model calibration for finite population parameter estimation under two stage sampling design. Communications in Statistics-Simulation and Computation, 1-17. https://doi.org/10.1080/03610918.2024.2369800

.

Sharma, M., Kumar, B., Mahajan, V. and Bhat, M.I.J. (2020). Ridge Regression Model for the Estimation of Total Carbon Sequestered by Forest Species. In: Chandra, G., Nautiyal, R., Chandra, H. (eds) Statistical Methods and Applications in Forestry and Environmental Sciences. Forum for Interdisciplinary Mathematics. Springer, Singapore. https://doi.org/10.1007/978981-15-1476-0_11

Valliant, R., Dorfman, A.H. and Royall, R.M. (2000). Finite population sampling and inference: a prediction approach. John Wiley, New York.

Wheeler, D.C. (2009). Simultaneous coefficient penalization and model selection in geographically weighted regression: the geographically weighted lasso. Environment and Planning A, 41(3), 722-742.