International Journal of Academic and Applied Research (IJAAR)   Year: 2023 | Volume: 7 | Issue: 2 | Page No.: 21-28 Comparison of Adaptive Bisquare And Adaptive Gaussian Weighting Functions In Poverty Case Modeling Using Geographically Weighted Regression In Papua Province Download PDF Aidawayati Rangkuti , Nurtiti Sunusi, Arya Winanda Tahir Abstract: Papua has the largest proportion of poor people in Indonesia. Predicting the poverty cases in Papua Province in 2021 using Geographically Weighted Regression (GWR) analysis, this study compared the adaptive bi-square and adaptive gaussian weighting functions. The SUSENAS data from the BPS's 2021 National Socio-Economic Survey were used. The analysis step that was conducted was using the OLS method. Using the test results, one significant variable was discovered. Additionally, the GWR approach was used for testing, and the values R2 and AIC of the GWR models with adaptive bi-square and adaptive gaussian weighting functions were compared. The value R2 and AIC of the GWR model with the bi-square adaptive weighting function were 94.5% and 147.0325, based on the findings. Using a gaussian adaptive weighting function, the GWR model's value R2 and AIC were 66.6% and 184.26. The GWR model with bi-square adaptive weighting function has the highest value R2 and the lowest AIC.