Dear Robin,
Many would vary on how best to do this.
The thoughts of many such people are given and/or linked to at
http://www.csse.monash.edu.au/~dld/mixture.modelling.page.html
My own preference for how best to do this is described at
http://www.csse.monash.edu.au/~dld/Snob.html
One little issue to add is that, since you have histograms, programs
which take into account the measurement accuracy (such as Snob) should
be given the width of the histogram(s) as the measurement accuracy.
Regards. - David Dowe.
http://www.csse.monash.edu.au/~dld/Snob.html
http://www.csse.monash.edu.au/~dld/mixture.modelling.page.html
> From owner-uai@ghost.CS.ORST.EDU Wed Mar 22 03:44:39 2000
> Subject: [UAI] Number of Gaussians in a histogram
> To: uai@CS.ORST.EDU
> Content-transfer-encoding: 7BIT
>
> Hi friends,
>
> Could you please let me know how to determine automatically
> the number of Gaussians present in a histogram. Since it is quite
> subjective, algorithms to determine the approximate number of
> Gaussians will be allright. Any pointer?
>
>
> Regards,
> robin
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com
This archive was generated by hypermail 2b29 : Tue Mar 21 2000 - 19:35:07 PST