commons-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Andreas Niekler <>
Subject [math] Usage of DifferentiableMultivariateRealFunction
Date Mon, 14 May 2012 08:02:29 GMT

i'm currently developing a gaussian process implementation for my PhD 
thesis. Within this package i need to calculate and maximize the 
marginal likelihood of the model given some data. This marginal 
likelihood is dependent on the parameters of a covariance function which 
is used. For a certain covariance function one can calculate the 
derivitives and the gradients for the marginal likelihood to optimize 
this function.

My Question: Can someone show me an example for the usage of 
DifferentiableMultivariateRealFunction and a matching optimizer. I 
understand the concept but the implementational details are hard to 
find. Here is what i understood:

value: the value of the function.

gradient(): an array containing all partial derivitivs gradients for the 

partialDerivitive: Value containing only one gradient of the function

if i provide this basic implementation i should pass this to the 
optimizer. Or do i have to provide an extra target function to the 
optimizer as inicated by some optinmizers with the <FUNC> operator?

Thnak you
Dipl.-Ing. (FH) Andreas Niekler
Mitarbeiter und Promovend
Bereich Multimedia-Produktionssysteme und -technologien

Hochschule für Technik, Wirtschaft und Kultur Leipzig
Fachbereich Medien

Gustav-Freytag-Straße 40
04277 Leipzig

Telefon: +49 0341 30 76 2378


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message