Brent [PK / PD]

posted by ElMaestro  – Denmark, 2013-07-18 11:07 (4302 d 09:20 ago) – Posting: # 11015
Views: 22,278

Hi d_labes,

❝ Can you enlighten me what the reason was to choose method="Brent"?


For the case of a one-dimensional optimisation problem, the Nelder-Mead is as unreliable as a used-car sales man. In such cases Brent is a good choice. If we want to do two-dimensional (or more) optimisation then I'd probably go with the default Nelder-Mead.

Along these lines: Yung-jin, for a two-dimensional problem as discussed above the Nelder-Mead etc. may perform much more efficiently than a grid search.

Re. the TTT rule: I have no experience with it in practice. It sounds like something that for all practical pruposes should be good enough.

Pass or fail!
ElMaestro

Complete thread:

UA Flag
Activity
 Admin contact
23,424 posts in 4,927 threads, 1,673 registered users;
95 visitors (0 registered, 95 guests [including 6 identified bots]).
Forum time: 20:28 CEST (Europe/Vienna)

There are two possible outcomes: if the result confirms the
hypothesis, then you’ve made a measurement. If the result is
contrary to the hypothesis, then you’ve made a discovery.    Enrico Fermi

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5