Uncertainty measuring and constraining method for geographic weighted regression model results

doi:10.11947/j.AGCS.2023.20210336

Abstract

Abstract: As a classical local weighting least-square method, the geographic weighted regression (GWR) model always suffers from the space sparsity of samples and the local multicollinearity of predictors, which results in the uncertainties of the model results show spatial heterogeneity. By constructing accuracy evaluation metric of posterior standard error based on the covariance propagation law, this study proposed an uncertainty measuring and constraining method for geographic weighted regression model and validated this method using the instance of ground PM_2.5 concentration remote sensing mapping. After uncertainty constraint, the results show the fitted accuracy and sample-based/site-based/regional-based cross validation accuracy for GWR model with different parameters are all improved; the sign error of regression coefficients caused by local multicollinearity are also corrected; the outlier and negative values in the GWR predicted values can also be effectively detected which improve the reliability of the ground PM_2.5 concentration mapping results. The proposed method can effectively guarantee the stability and effectiveness of GWR results.

Key words: geographic weighted regression, uncertainty measuring, PM_2.5 concentration, remote sensing mapping

CLC Number:

P227

LIU Ning, ZOU Bin, ZHANG Honghui. Uncertainty measuring and constraining method for geographic weighted regression model results[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(2): 307-317.

References

[1] BRUNSDON C, FOTHERINGHAM S, CHARLTON M. Geographically weighted regression[J]. Journal of the Royal Statistical Society:Series D (the Statistician), 1998, 47(3):431-443.
[2] HURVICH C M, SIMONOFF J S, TSAI C L. Smoothing parameter selection in nonparametric regression using an improved Akaike information criterion[J]. Journal of the Royal Statistical Society:Series B (Statistical Methodology), 1998, 60(2):271-293.
[3] BOWMAN A W. An alternative method of cross-validation for the smoothing of density estimates[J]. Biometrika, 1984, 71(2):353-360.
[4] MA Zongwei, HU Xuefei, HUANG Lei, et al. Estimating ground-level PM_2.5 in China using satellite remote sensing[J]. Environmental Science & Technology, 2014, 48(13):7436-7444.
[5] 汤庆园, 徐伟, 艾福利. 基于地理加权回归的上海市房价空间分异及其影响因子研究[J]. 经济地理, 2012, 32(2):52-58. TANG Qingyuan, XU Wei, AI Fuli. A GWR-based study on spatial pattern and structural determinants of Shanghai's housing price[J]. Economic Geography, 2012, 32(2):52-58.
[6] DUAN Sibo, LI Zhaoliang. Spatial downscaling of MODIS land surface temperatures using geographically weighted regression:case study in Northern China[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(11):6458-6469.
[7] 卢宾宾, 葛咏, 秦昆, 等. 地理加权回归分析技术综述[J]. 武汉大学学报(信息科学版), 2020, 45(9):1356-1366. LU Binbin, GE Yong, QIN Kun, et al. A review on geographically weighted regression[J]. Geomatics and Information Science of Wuhan University, 2020, 45(9):1356-1366.
[8] 胡炜, 刘永学, 林勇军. 基于GWR的住宅地价相对修正方法研究:以深圳市为例[J]. 中国土地科学, 2017, 31(9):62-69, 2, 97. HU Wei, LIU Yongxue, LIN Yongjun. Study on the correcting method of residential land price based on GWR:a case study of Shenzhen city[J]. China Land Sciences, 2017, 31(9):62-69, 2, 97.
[9] WHEELER D, TIEFELSDORF M. Multicollinearity and correlation among local regression coefficients in geographically weighted regression[J]. Journal of Geographical Systems, 2005, 7(2):161-187.
[10] SAMKAR H, ALPU O. Ridge regression based on some robust estimators[J]. Journal of Modern Applied Statistical Methods, 2010, 9(2):495-501.
[11] GUJARATI D, PORTER D. Basic econometrics[M]. New York:McGraw-Hill Education, 2008.
[12] LLOYD C D. Analysing population characteristics using geographically weighted principal components analysis:a case study of Northern Ireland in 2001[J]. Computers, Environment and Urban Systems, 2010, 34(5):389-399.
[13] STEWART F A, YANG W, KANG W. Multiscale geographically weighted regression (MGWR)[J]. Annals of the American Association of Geographers, 2017, 107(6):1247-1265.
[14] YU H, STEWART F A, LI Z, et al. On the measurement of bias in geographically weighted regression models[J]. Spatial Statistics, 2020, 38:100453.
[15] POURMOHAMMADI P, STRAGER M P, DOUGHERTY M J, et al. Analysis of land development drivers using geographically weighted ridge regression[J]. Remote Sensing, 2021, 13(7):1307.
[16] BELSLEY D A, KUH E, WELSCH R E. Regression diagnostics:identifying influential data and sources of collinearity[M]. New York:John Wiley & Sons, 1980.
[17] FOX J, MONETTE G. Generalized collinearity diagnostics[J]. Journal of the American Statistical Association, 1992, 87(417):178-183.
[18] JEUDY L M A. Generalyzed variance-covariance propagation law formulae and application to explicit least-squares adjustments[J]. Bulletin Géodésique, 1988, 62(2):113-124.
[19] SONG Yingchun, XIA Yuguo, XIE Xuemei. Adjustment model and algorithm based on ellipsoid uncertainty[J]. Journal of Geodesy and Geoinformation Science, 2020, 3(3):59-66.
[20] OSHAN T, LI Z, KANG W, et al.MGWR:a python implementation of multiscale geographically weighted regression for investigating process spatial heterogeneity and scale[J]. ISPRS International Journal of Geo-Information, 2019, 8(6):269.
[21] YU H, STEWART F A, LI Z, et al. Inference in multiscale geographically weighted regression[J]. Geographical Analysis, 2020, 52(1):87-106.
[22] RODRIGUEZ J D, PEREZ A, LOZANO J A. Sensitivity analysis of k-fold cross validation in prediction error estimation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(3):569-575.
[23] 归庆明, 姚绍文, 顾勇为, 等. 诊断复共线性的条件指标-方差分解比法[J]. 测绘学报, 2006,35(3):210-214. GUI Qingming, YAO Shaowen, GU Yongwei, et al. A new method to diagnose multicollinearity based on condition index and variance decomposition proportion(CIVDP)[J]. Acta Geodaetica et Cartographica Sinica, 2006, 35(3):210-214.
[24] LI Tongwen, SHEN Huanfeng, ZENG Chao, et al. A validation approach considering the uneven distribution of ground stations for satellite-based PM_2.5 estimation[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13:1312-1321.
[25] WANG Jun. Intercomparison between satellite-derived aerosol optical thickness and PM_2.5 mass:implications for air quality studies[J]. Geophysical Research Letters, 2003, 30(21):2095.
[26] ZHANG Tianhao, GONG Wei, WANG Wei, et al. Ground level PM_2.5 estimates over China using satellite-based geographically weighted regression (GWR) models are improved by including NO₂ and enhanced vegetation index (EVI)[J]. International Journal of Environmental Research and Public Health, 2016, 13(12):1215.