%0 Book Section
%9 OS CH : Chapitres d'ouvrages scientifiques
%A Ndiaye, M.
%A Dabo-Niang, S.
%A Ngom, P.
%A Thiam, N.
%A Brehmer, Patrice
%A El Vally, Y.
%T Nonparametric prediction and supervised classification for spatial dependent functional data under fixed sampling design
%B Nonlinear analysis, geometry and applications
%C Cham
%D 2024
%E Seck, D.
%E Kangni, K.
%E Salomon Sambou, M.
%E Nang, P.
%E Fall, M.M.
%L fdi:010093033
%G ENG
%I Birkhäuser
%@ 978-3-031-52680-0
%K SENEGAL ; ATLANTIQUE
%P 69-100
%R 10.1007/978-3-031-52681-7_3
%U https://www.documentation.ird.fr/hor/fdi:010093033
%> https://www.documentation.ird.fr/intranet/publi/2025-02/010093033.pdf
%W Horizon (IRD)
%X Fisheries science has been trying to identify the best way to analyze and predict fish biomass and its spatial distribution since several decades using, among others, kirigng model, co-kiriging model, Species Distribution Modeling and Joined Species Distribution Modeling, based on conventional statistical methods as Generalized Linear Models and Generalized Additive Models, with contested results. We consider a bio-ecological issue applying a non parametric spatial prediction based on a spatio-functional regression models, in a fixed design sampling context, as a supervised classification method when the variable of interest belongs to a predefined class set. The proposed predictor takes into account the spatial fish distribution and environmental variable such as salinity and temperature. The development of the method depends on two kernels to control both interactions between observations and locations. The results show that this nonparametric spatial functional supervised classification method is an efficient tool applied to predict spatial distribution of demersal coastal fish off Senegal.
%S Trends in Mathematics
%B The NLAGA's Biennial International Research Symposium
%8 2023/08/21-27
%$ 126TELAPP05 ; 126TELTRN05