Spatial Function of Influence on Center Optimal Location Based on L_p-Norms

摘要

We propose a sensitivity analysis using generalized L_p-norm (Minkowski distance) applied on center optimal location (1 facility). The results show that there exists in one dimension an underlying (log)linear relation between influence and distance of the demand points on the center. New L_p-norms are emphasized with interesting properties in statistics (e.g. with p=3) although they are not used in location optimization. The law we enhance is of interest in both statistics and and spatial analysis domains and highlights in a new way the impact of the metrics choice on the center location, through the induced spatial influence function, those metrics aiming at spatial equity (L_∞), equality (L_2) or efficiency (L_1).