[PS05] PLS discriminant analysis for functional data
Conférence Internationale avec comité de lecture :
ASMDA'05 XIth Int. Symp. on Applied Stochastic Models and Data Analysis, Brest,
January 2005,
pp.653-661,
motcle:
Résumé:
Partial least squares regression on functional data is applied in the
context of linear discriminant analysis with binary response. The discriminant coefficient function is then used to compute score to one of the two classes. The method is applied to gait data and the results are compared with those given by linear discriminant analysis and logistic regression on the principal components of predictors.