Dear experts,
I've got a review to a paper where one of the reviewers asks us to analyze the data using a "general additive model" (GAD) instead of the usual GLM.
I've done a bit of reading about GAD and I don't get them, at least in the most general sense. The wikipedia page says that GAD model the data as y = c + sum_i b_i * f_i (x) instead of y = c + sum_i b_i *x_i. This definition is way too vague to be useful, unless one has an idea regarding what the f(x) should look like (and, for situations like neuroimaging, they seem way too prone to overfitting).
So I was just wondering whether one of you ever used GAD in the field of neuroimaging, whether is there some package around (SPM compatible maybe? :D) and whether is there any useful reference around (haven't been able to find them).
Thank you a lot,
Luca
Rispetta l’ambiente: non stampare questa mail se non è necessario.
Respect the environment: if it's not necessary, don't print this mail.
A NATALE FAI UN REGALO ALLA RICERCA
Questo Natale scegli di fare il dono più bello: un regalo al progresso della ricerca.
Scopri come su http://www.hsr.it/sostienici/regali-solidali <http://www.hsr.it/sostienici/regali-solidali/> oppure DONA SUBITO online su http://register.sanraffaele.org/donazioni/donazione.aspx .
Non c’è cura, senza ricerca.
[http://dsi.hsr.it/static/banner.jpg]
|