We define a class of SNR models for non-grouped data as

where 's are responses; is a covariate in a general domain ; is a known function of which depends on a vector of parameters and a vector of unknown non-parametric function ; and are random errors distributed as . 's are modeled using SS ANOVA models as represented in (). We assume that depends on a parsimonious set of parameters .

For model (), we denote
and
.
Model () can then be written in the vector form

Yuedong Wang 2004-05-19