Hi again ecologgers I presume that in looking for the "best" model in a CART you are looking for the lowest residual mean deviance (besides the biological meaning) ?? or to put it in another perspective, what is the most sensible procedure to arrive a the most stable solution without getting into overfitting.
<P>Abraham de Alba Avila</P> <P>Terrestrial Plant Ecology</P> <P>INIFAP-Ags</P> <P> Ap. postal 20,</P> <P> Pabellón Arteaga, 20660</P> <P> Aguascalientes, MEXICO</P> <P> Tel: (465) 95-801-67, & 801-86 ext. 118, FAX ext 102 alternate: [EMAIL PROTECTED] cel: 449-157-7070</P> __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com