Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hello from Edmonton (and on Temporal Differences)

Author: Miguel A. Ballicora

Date: 19:24:45 08/05/02

Go up one level in this thread


On August 05, 2002 at 18:13:33, Dieter Buerssner wrote:

>On August 04, 2002 at 15:43:49, Miguel A. Ballicora wrote:
>
>>5000 parameters is not much when compared to the parameters needed to obtain the
>>optimal conformation of a protein with a computer. In that case, it is almost
>>impossible (with the current knowledge) to obtain the right conformation
>>starting from scratch, but is is very doable when you start from the "optimal"
>>conformation (determine by physical methods, not by a computer), you change
>>something and see how the new conformation would look like. Iterations around
>>the minimum are very fast because all the parameters behave almost linearly or
>>close enough. When the parameters behave linearly the time to resolve the
>>problem is O(1) (a linear regresion of n parameters is O(1)).
>
>Miguel, I cannot follow here. It is some time ago, but I studied many
>multi-parameter optimization methods. My memory tells me, that for linear

Sorry, I misused the O() notation, my mistake. I was talking about iterations.
A liner regression takes 1 iteration or cycle. The more linear the parameters
behave, the less iterations you need, even for approximations (those do not need
to be O(n^3)).

Regards,
Miguel



>regression (the easiest case - not?) this involves solving a linear system of n
>equations. This would be O(n^3) - certainly not a trivial task for n=5000,
>besides all the (serious) numerical instabilities, one could await for this task
>- especially when trying to solve it in a rather straightforward manner.
>
>Regards,
>Dieter



This page took 0.01 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.