A Possibilistic Regression Based on Gradual Interval B-splines: Application for hyperspectral imaging Lake Sediments

Abstract According to an epistemic view of intervals, this paper proposes a possibilistic regression based on gradual interval B-splines to represent the input-output data mapping. In this context, an improvement of the parametric fuzzy regression through the notion of gradual interval B-splines is proposed. The proposed gradual regression based on B-splines can be regarded as an extension of the nonparametric interval-based regression where an uncertain dimension is integrated. The proposed method is validated through illustrative and comparative examples. Moreover, it is applied for modelling the input-output behavior between reflectance and wavelengths in an application for hyperspectral imaging for lake sediments.