[Taxacom] methodological plurality
dlahr at ib.usp.br
Thu Nov 22 12:34:19 CST 2012
> What priors should be used for phylogeny is an interesting issue I would
> like to see further explored.
I have only seem some attempts at this that seemed to make sense by Mark
Pagel. He uses a likelihood surface and some other smart guesses to try
and narrow down where the possible answer should be. In other words, he
gets a "feel" for the data by exploring it using likelihood methods, and
then moves to a BI framework.
Seems interesting, but also seems circular - use likelihood to estimate
parameters then use estimated parameters as priors for likelihood
analysis. But I may be wrong my math skills don´t go so far.
I honestly have never seen anything more palpable than this, but would love
Daniel J. G. Lahr, PhD
Dept of Zoology, Univ. of Sao Paulo, Brazil
More information about the Taxacom