-------- Original Message -------- Subject: Re: more discriminant procedures Date: Sat, 18 Oct 2008 11:27:42 +0200 From: [EMAIL PROTECTED] To: morphmet <[EMAIL PROTECTED]> References: <[EMAIL PROTECTED]> A good article is that of N A Campbell and Atchley W R (1981). The geometry of canonical variate analysis. Syst Zool, 30, 268-280. A well reasoned study in which the importance of stability is presented in accessible terms in relation to the reification of canonical vectors. A useful text for the tyro is the little book by Sir Maurice Kendall (1957) A Course in Multivariate Analysis. 185 pp, Griffin, London Old but still very good. (Several times reprinted.) Historical Note Still also very good is the book in which the basics of CVA and other multivariate methods were presented in exemplarily lucid form: C. Radakrishna Rao (1952). Advanced Statistical methods in Biometric Research. 292 pp. Wiley, NY Rao is responsible for tidying up the state of canonical variates (multiple discrimination) and the generalized statistical distance as part of his thesis work in the UK for Professor Maurice Bartlett in 1946 (Published in the Journal of the Royal Statistical Society, London). Sir Ronald Aylmer Fisher certainly had the first notions (additionally also several of the brillian statisticians of the Indian Institute of Statistics, Calcutta under the guidance of P. C. Mahanalobis, for whom the generalized statistical distance was eventually named). All you need to know about discriminant functions is contained in the NATO publication edited by Professor T. Cacoullos (1973) entitled Discriminant Analysis and Applications, 434 pp. Academic Press NY Richard A. Reyment [EMAIL PROTECTED] ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. -- Replies will be sent to the list. For more information visit http://www.morphometrics.org
