Investigating the harmonization of highly noisy heterogeneous datasets hand-collected over the same study domain

Résumé fourni par la source

The objective of this paper is to propose an approach to harmonise noisy spatial data acquired by different operators using (low-cost) hand-held sensors over the same spatial domain. In such cases, datasets need to be harmonised first to be comparable before decision making. This work proposes a methodology to address this issue in the case of nested and noisy spatial data. First, it proposes the implementation of a non-parametric test of Kolmogorov-Smirnov to determine if harmonisation is needed. Then, it proposes an aspatial harmonization method based on standardization. The method was applied on grape sugar content datasets collected by 2 hand-held spectrometers. Results showed that harmonizing a less confident dataset is interesting solely if the size of the trusted one is too small.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.