[Date Index] [Thread Index] [Author Index]

Re: Maximization of a Likelihood Function

• To: mathgroup at smc.vnet.net
• Subject: [mg91483] Re: Maximization of a Likelihood Function
• From: Bill Rowe <readnews at sbcglobal.net>
• Date: Sat, 23 Aug 2008 01:42:46 -0400 (EDT)

On 8/22/08 at 3:10 AM, gregory.lypny at videotron.ca (Gregory Lypny)
wrote:

>Just curious to know if anyone has written a Mathematica routine to
>maximize a likelihood function.  I'd like to tinker with one and
>apply it to multivariate GARCH if that is possible.  I know that the
>Time Series package has a univariate GARCH command, but I am unaware
>of any multivariate routines.

There is no need to write a routine to maximize a likelihood
function. The built-in commands, Maximize or NMaximize will
handle multiple dimensions. All that is needed is a correctly
formulated likelihood function and to use correct syntax with
either Maximize or NMaximize.