@article{fdi:010086044,
  title = {{R}esponse mixture models based on supervised components : clustering floristic taxa},
  author = {{G}ibaud, {J}. and {B}ry, {X}. and {T}rottier, {C}. and {M}ortier, {F}. and {R}{\'e}jou-{M}{\'e}chain, {M}axime},
  editor = {},
  language = {{ENG}},
  abstract = {{I}n this article, we propose to cluster responses in order to identify groups predicted by specific explanatory components. {A} response matrix is assumed to depend on a set of explanatory variables and a set of additional covariates. {E}xplanatory variables are supposed many and redundant, which implies some dimension reduction and regularization. {B}y contrast, additional covariates contain few selected variables which are forced into the regression model, as they demand no regularization. {T}he response matrix is assumed partitioned into several unknown groups of responses. {W}e suppose that the responses in each group are predictable from an appropriate number of specific orthogonal supervised components of explanatory variables. {T}he classification is based on a mixture model of the responses. {T}o estimate the model, we propose a criterion extending that of {S}upervised {C}omponent-based {G}eneralized {L}inear {R}egression, a {P}artial {L}east {S}quares-type method, and develop an algorithm combining component-based model and {E}xpectation {M}aximization estimation. {T}his new methodology is tested on simulated data and then applied to a floristic ecology dataset.},
  keywords = {{EM} algorithm ; {R}esponse mixture ; {SCGLR} ; {S}upervised components ; {T}axa classification},
  booktitle = {},
  journal = {{S}tatistical {M}odelling},
  volume = {[{E}arly access]},
  numero = {},
  pages = {[19 ]},
  ISSN = {1471-082{X}},
  year = {2022},
  DOI = {10.1177/1471082x221115525},
  URL = {https://www.documentation.ird.fr/hor/fdi:010086044},
}