Une adaptation de la régression bidimensionnelle aux polygones : un exemple en cognition spatiale

Bidimensional regression is a statistical method used to compare two surfaces. It is based on the adjustment of the positions of two clouds of homologous points structuring the two surfaces to compare. The results of the bidimensional regression only concern the clouds of points in question. The article presents a method to extend the results to polygons, especially in a mapping context. For this purpose, it complements the bidimensional regression by two successive interpolations. To illustrate its possibilities, the method is applied to the mapping of cognitive representations of forest fragments in Thiérache (France).