Skip to main content

Table 2 Comparison of bandwidth types, kernel shapes and bandwidth optimization methods

From: Do the risk factors for type 2 diabetes mellitus vary by location? A spatial analysis of health insurance claims in Northeastern Germany using kernel density estimation and geographically weighted regression

Modell (bandwidth type, kernel shape, optimization method) AICc Adjusted R2 Moran’s I of residuals
Adaptive, Gaussian, AICc −347 0.51 p < 0.001
Adaptive, Gaussian, AIC −347 0.51 p < 0.001
Adaptive, Gaussian, BIC −315 0.44 p < 0.001
Adaptive, Gaussian, CV −347 0.51 p < 0.001
Fixed, Gaussian, AICc −385 0.62 p < 0.05
Fixed, Gaussian, AIC −265 0.66 p > 0.05
Fixed, Gaussian, BIC −316 0.44 p < 0.001
Fixed, Gaussian, CV −370 0.64 p > 0.05
Adaptive, bi-square, AICc −394 0.63 p < 0.001
Adaptive, bi-square, AIC −374 0.66 p > 0.05
Adaptive, bi-square, BIC −320 0.45 p < 0.001
Fixed, bi-square, AICc −385 0.62 p < 0.01
Fixed, bi-square, AIC 40 0.68 p > 0.05
Fixed, bi-square, BIC −316 0.44 p < 0.001