commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ajo Fod (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (MATH-1092) NonLinearConjugateGradientOptimizer's Line search is a gradient search returns obviously suboptimal point.
Date Wed, 22 Jan 2014 15:49:19 GMT

     [ https://issues.apache.org/jira/browse/MATH-1092?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Ajo Fod updated MATH-1092:
--------------------------

    Description: 
In package : org.apache.commons.math3.optim.nonlinear.scalar.gradient

In a minimization problem, a line search should not return a point where the value is greater
than the values at the edges of the interval. The line search violates this obvious requirement
by focusing solely on solving for gradient=0 and ignoring the value.

Moreover why not use the BrentOptimizer to do the line search? Its a reasonably well tested
piece of code that does just what LinesearchFunction is supposed to do.





  was:
In package : org.apache.commons.math3.optim.nonlinear.scalar.gradient

In a minimization problem, a line search should not return a point where the value is greater
than the values at the edges of the interval. The line search violates this obvious requirement
by focusing solely on solving for gradient=0 and ignoring the value.

Moreover why not use the BrentOptimizer? Its a reasonably well tested piece of code that does
just what LinesearchFunction is supposed to do.






> NonLinearConjugateGradientOptimizer's Line search is a gradient search returns obviously
suboptimal point.
> ----------------------------------------------------------------------------------------------------------
>
>                 Key: MATH-1092
>                 URL: https://issues.apache.org/jira/browse/MATH-1092
>             Project: Commons Math
>          Issue Type: Bug
>            Reporter: Ajo Fod
>
> In package : org.apache.commons.math3.optim.nonlinear.scalar.gradient
> In a minimization problem, a line search should not return a point where the value is
greater than the values at the edges of the interval. The line search violates this obvious
requirement by focusing solely on solving for gradient=0 and ignoring the value.
> Moreover why not use the BrentOptimizer to do the line search? Its a reasonably well
tested piece of code that does just what LinesearchFunction is supposed to do.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)

Mime
View raw message