A robust geographically weighted LASSO model with Huber Loss for outlier-resilient spatial analysis: application to COVID-19 dynamics in Indian districts
Main Article Content
Abstract
Geographically Weighted Regression (GWR) is a widely used technique for modeling spatially varying relationships in epidemiological data. However, its sensitivity to multicollinearity and outliers can limit its reliability in complex datasets, such as those involving COVID-19. To address these limitations, this study first employed the RGWR model, which uses the Huber loss to enhance robustness against outliers. Simultaneously, to handle multicollinearity and enable local variable selection, the Geographically Weighted LASSO (GWL) model is employed. Despite their individual strengths, both RGWR and GWL face challenges when outliers and multicollinearity coexist. To overcome this, we propose a Robust Geographically Weighted LASSO (RGWL) model that integrates the Huber loss function with a spatially adaptive LASSOpenalty. This combined formulation enhances robustness to outliers while effectively performing localized regularization and variable selection. We apply these models to district-level COVID-19 data from India, incorporating a comprehensive set of socio-demographic, environmental, and health-related variables. Comparative analysis reveals that the proposed RGWL model outperforms both RGWR and GWL in terms of predictive accuracy and goodness of fit, particularly in the presence of extreme observations and collinear predictors. The spatially varying coefficients further highlight region-specific patterns, demonstrating the utility of the RGWL model for robust, localized analysis in epidemiological research.
Article Details
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
References
1. Adekunle, I. A., Onanuga, A. T., Akinola, O. O. & Ogunbanjo, O. W. Modelling spatial
variations of coronavirus disease (COVID-19) in Africa. Science of the Total Environment 729,
138998. doi:10.1016/j.scitotenv.2020.138998 (2020).
2. Ahmed, F., Ahmed, N., Pissarides, C. & Stiglitz, J. Why inequality could spread COVID-19.
The Lancet Public Health 5, e240. doi:10.1016/S2468-2667(20)30085-2 (2020).
3. Alfons, A., Croux, C. & Gelper, S. Sparse least trimmed squares regression for analyzing highdimensional large data sets. Annals of Applied Statistics 7, 226–248. doi:10.1214/12-AOAS575
(2013).
4. Appiah-Otoo, I. & Kursah, M. B. Modelling spatial variations of novel coronavirus disease
(COVID-19): evidence from a global perspective. GeoJournal 87, 3203–3217. doi:10 . 1007 /
s10708-021-10427-0 (2022).
5. Barros, A. C. A. V. B. d., Silva, M. A., Marques, R. D., Lobos, C. M. V. & Luciano, A. C. d. S.
Classification of sugarcane areas in Landsat images using machine learning algorithms. Brazilian
Journal of Biometrics 42, 147–157. doi:10.28951/bjb.v42i2.693 (2024).
6. Bashir, M. F., Ma, B., Komal, B., Bashir, M. A., Tan, D., Bashir, M., et al. Correlation between
climate indicators and COVID-19 pandemic in New York, USA. Science of the Total
Environment 728, 138835. doi:10.1016/j.scitotenv.2020.138835 (2020).
7. Brunsdon, C., Fotheringham, S. & Charlton, M. Geographically weighted regression. Journal
of the Royal Statistical Society: Series D (The Statistician) 47, 431–443. doi:10 . 1111 / j . 1538 -
4632.1996.tb00936.x (1998).
8. CARMO, C. R. S., de Lima CANEPPELE, F. & NUNES, F. C. Analysis of covid-19 contamination
and deaths cases in brazil according to the newcomb-benford. Brazilian Journal of
Biometrics 39, 522–535. doi:10.28951/rbb.v39i4.535 (2021).
9. Carrilho, J. & Lopes, M. B. Classification and biomarker selection in lower-grade glioma using
robust sparse logistic regression applied to RNA-seq data. Brazilian Journal of Biometrics 40,
371–381. doi:10.28951/bjb.v40i4.634 (2022).
10. Chen, B., Liang, H., Yuan, X., Hu, Y., Xu, M., Zhao, Y., Zhang, B., Tian, F. & Zhu, X.
Roles of meteorological conditions in COVID-19 transmission on a worldwide scale. MedRxiv,
2020–03. doi:10.1101/2020.03.16.20037168 (2020).
11. Chung, H. W. et al. Effects of government policies on the spread of COVID-19 worldwide.
Scientific Reports 11, 20495. doi:10.1038/s41598-021-99368-9 (2021).
12. Cliff, A. D. & Ord, J. K. Spatial Processes: Models & Applications (Pion, London, 1981).
13. Dutta, I., Basu, T. & Das, A. Spatial analysis of COVID-19 incidence and its determinants
using spatial modeling: A study on India. Environmental Challenges 4, 100096. doi:10.1016/j.
envc.2021.100096 (2021).
14. Fan, J., Fan, Y. & Barut, E. Adaptive robust variable selection. Annals of Statistics 40, 1748–
1772. doi:10.1214/13-AOS1191 (2012).
15. Foody, G. M. Local characterization of thematic classification accuracy through spatially constrained confusion matrices. International Journal of Remote Sensing 24, 1217–1225. doi:10.1080/01431160512331326521 (2003).
16. Gujarati, D. N. Basic Econometrics 4th (McGraw-Hill, New York, 2004).
17. Khan, M., Javed, M. & Ahmed, S. A robust penalized regression estimator for simultaneous
variable selection and outlier detection. Statistical Papers 63, 1151–1173. doi:10.1007/s00362-
021-01304-7 (2022).
18. NEISSE, A. C., de OLIVEIRA, F. L. P., de OLIVEIRA, A. C. S., NETO, R. M. N., et al.
Chronic fatigue syndrome and its relation with absenteeism: elastic-net and stepwise applied
to biochemical and anthropometric clinical measurements. Brazilian Journal of Biometrics 39,
221–239. doi:10.28951/rbb.v39i1.533 (2021).
19. Pereira, M. & Oliveira, A. M. Poverty and food insecurity may increase as the threat of
COVID-19 spreads. Public health nutrition 23, 3236–3240. doi:10.1017/S1368980020003493
(2020).
20. Ramírez-Aldana, R., Gomez-Verjan, J. C. & Bello-Chavolla, O. Y. Spatial analysis of COVID-
19 spread in Iran: Insights into geographical and structural transmission determinants at a
province level. PLoS neglected tropical diseases 14, e0008875. doi:10.1371/journal.pntd.0008875
(2020).
21. Rocklöv, J. & Sjödin, H. High population densities catalyse the spread of COVID-19. Journal
of travel medicine 27, taaa038. doi:10.1093/jtm/taaa038 (2020).
22. Sardar, S., Abdul-Khaliq, I., Ingar, A., Amaidia, H. & Mansour, N. COVID-19 lockdown: A
protective measure or exacerbator of health inequalities? A comparison between the United
Kingdom and India.’a commentary on “the socio-economic implications of the coronavirus
and COVID-19 pandemic: A review. International Journal of Surgery (London, England) 83, 189.
doi:10.1016/j.ijsu.2020.09.044 (2020).
23. Sarkar, S. K., Ekram, K. M. M. & Das, P. C. Spatial modeling of COVID-19 transmission in
Bangladesh. Spatial Information Research, 1–12. doi:10.1007/s41324-021-00387-5 (2021).
24. Sridhar, K. S. Urbanization and COVID-19 Prevalence in India. Regional Science Policy &
Practice 15, 493–505. doi:10.1111/rsp3.12503 (2023).
25. Tosepu, R., Gunawan, J., Effendy, D. S., Lestari, H., Bahar, H., Asfian, P., et al. Correlation
between weather and Covid-19 pandemic in Jakarta, Indonesia. Science of the total environment
725, 138436. doi:10.1016/j.scitotenv.2020.138436 (2020).
26. Wang, H., Li, G. & Jiang, G. Robust regression shrinkage and consistent variable selection
through the LAD-Lasso. Journal of Business & Economic Statistics 25, 347–355. doi:10.1198/
073500106000000251 (2007).
27. Wheeler, D. C. Diagnostic tools and a remedial method for collinearity in geographically
weighted regression. Environment and planning A 39, 2464–2481. doi:10.1068/a38325 (2007).
28. Wheeler, D. C. Simultaneous coefficient penalization and model selection in geographically
weighted regression: the geographically weighted lasso. Environment and planning A 41, 722–
742. doi:10.1068/a40256 (2009).
29. Wulandari, P. P., Djuraidah, A. & Wigena, A. H. Robust Geographically Weighted Regression
Modeling using Least Absolute Deviation and M-Estimator. International Journal of Scientific
Research in Science, Engineering and Technology 6, 238–245. doi:10.32628/IJSRSET196123
(2019).
30. Zhang, H. & Mei, C. Local least absolute deviation estimation of spatially varying coefficient
models: robust geographically weighted regression approaches. International Journal of Geographical Information Science 25, 1467–1489. doi:10.1080/13658816.2010.528420 (2011).