Talk:Nonlinear conjugate gradient method
Add topic | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
Please define the variable N. (Is it the number of dimensions of x?) It is used in the text with out definition.
- The reference to N has been removed. Perhaps someone was mistaking this page for the linear conjugate gradient method.lgstarn (talk) 17:05, 17 November 2008 (UTC)
The examples for choosing Beta are not a matter of taste. As described, only FR guarentees to converge. However, in practice it is observed to converge slowly. PR may no converge, however when it does it is much faster. See Nocedal and Wright. —Preceding unsigned comment added by 76.164.46.79 (talk) 21:03, 6 January 2010 (UTC)
Adding re-directs
[edit]Could someone add a redirect from Non-linear conjugate gradient method? --134.102.204.124 (talk) 11:08, 23 September 2010 (UTC)
Problem definition
[edit]I find it strange to immediately specialize the quadratic objective to the special case of least-squares: it's nonstandard, potentially confusing, and unnecessary, in my mind. Maybe I'm missing the rationale? — Preceding unsigned comment added by 154.5.53.184 (talk) 21:14, 1 June 2012 (UTC)