Dont worry, I have got Minka's paper
Thanks
XC
****************************************
Xuelin Cui
Department of Electrical Engineering
University of Hawaii-Manoa
Honolulu HI 96822
Tel: 1-808-349-0983
Email: [log in to unmask]
****************************************
----- Original Message -----
From: Christian Beckmann <[log in to unmask]>
Date: Wednesday, January 31, 2007 11:08 am
Subject: Re: [FSL] how to determine the number of independent components?
To: [log in to unmask]
> That's right, this is the technical report I'm referring to (as
> opposed to the paper published in IEEE TMI with the same title but
> different formating, e.g. different page numbering and section
> numbering). Section 3.1. reviews Thomas Minka's technical report
> which introduced the approximations to the full model evidence in
> the
> context of Probabilistic PCA (see ref 37 in our technical report).
> ta
> christian
>
>
>
> On 31 Jan 2007, at 20:58, Xuelin Cui wrote:
>
> > Christian:
> >
> > I check the link again which is: http://www.fmrib.ox.ac.uk/
> > analysis/techrep/tr02cb1/tr02cb1.pdf
> > This paper looks same with the paper with the title that I sent
> to
> > you in the previous Email. If not this paper, could you please
> > send me the exact link again that introduces how to determine the
>
> > number of ICs exactly?
> >
> > Thanks a lot, and very sorry to take your time
> >
> > Xuelin
> >
> > ****************************************
> > Xuelin Cui
> > Department of Electrical Engineering
> > University of Hawaii-Manoa
> > Honolulu HI 96822
> >
> > Tel: 1-808-349-0983
> > Email: [log in to unmask]
> > ****************************************
> >
> > ----- Original Message -----
> > From: Christian Beckmann <[log in to unmask]>
> > Date: Wednesday, January 31, 2007 10:46 am
> > Subject: Re: [FSL] how to determine the number of independent
> > components?
> > To: [log in to unmask]
> >
> >> Hi
> >>
> >> no, by technical report I mean the technical report, the link was
> >> contained in one of my earlier email (Jan 29th I believe). You can
> >> read the TMI paper if you want, the eqn and page numbers will be
> >> different, though.
> >> ta
> >> christian
> >>
> >>
> >>
> >> On 31 Jan 2007, at 20:29, Xuelin Cui wrote:
> >>
> >>> hi:
> >>> Thanks Christian.
> >>> So what do you mean by techinical report is the paper of
> >>> ""Probabilistic Independent Component Analysis for Functional
> >>> Magnetic Resonance Imaging"? Thus, in its section 3.1, the way of
> >>
> >>> determine the number of ICs is introduced?
> >>>
> >>> Thanks
> >>>
> >>> Xuelin
> >>>
> >>> ****************************************
> >>> Xuelin Cui
> >>> Department of Electrical Engineering
> >>> University of Hawaii-Manoa
> >>> Honolulu HI 96822
> >>>
> >>> Tel: 1-808-349-0983
> >>> Email: [log in to unmask]
> >>> ****************************************
> >>>
> >>> ----- Original Message -----
> >>> From: Christian Beckmann <[log in to unmask]>
> >>> Date: Wednesday, January 31, 2007 10:15 am
> >>> Subject: Re: [FSL] how to determine the number of independent
> >>> components?
> >>> To: [log in to unmask]
> >>>
> >>>> Hi
> >>>>
> >>>> melodic estimates the number of components by calculating the
> >>>> approximate model evidence for each possible number following eqn
> >>>> 10
> >>>> in the technical report. Section 3.1 explains this
> >>>> thanks
> >>>> christian
> >>>>
> >>>>
> >>>> On 31 Jan 2007, at 20:08, Xuelin Cui wrote:
> >>>>
> >>>>> hi folks:
> >>>>>
> >>>>> I ran MELODIC, and get some results. But I dont really know how
> >>>>> MELODIC gets the ICs. How does it determine the number of
> >>>>> ICs? Is the number of ICs equal to the total time points? Where
> >>>> can
> >>>>> I see the ICA algorithm that MELODIC exactly follows?
> >>>>>
> >>>>> Thanks a lot
> >>>>>
> >>>>> Xuelin
> >>>>>
> >>>>> ****************************************
> >>>>> Xuelin Cui
> >>>>> Department of Electrical Engineering
> >>>>> University of Hawaii-Manoa
> >>>>> Honolulu HI 96822
> >>>>>
> >>>>> Tel: 1-808-349-0983
> >>>>> Email: [log in to unmask]
> >>>>> ****************************************
> >>>>
> >>>> --
> >>>> Christian F. Beckmann
> >>>> Oxford University Centre for Functional
> >>>> Magnetic Resonance Imaging of the Brain,
> >>>> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> >>>> Email: [log in to unmask] -
> >>>> http://www.fmrib.ox.ac.uk/~beckmann/ Phone: +44(0)1865 222551
> Fax:>>>> +44(0)1865 222717
> >>>>
> >>
> >> --
> >> Christian F. Beckmann
> >> Oxford University Centre for Functional
> >> Magnetic Resonance Imaging of the Brain,
> >> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> >> Email: [log in to unmask] -
> >> http://www.fmrib.ox.ac.uk/~beckmann/ Phone: +44(0)1865 222551 Fax:
> >> +44(0)1865 222717
> >>
>
> --
> Christian F. Beckmann
> Oxford University Centre for Functional
> Magnetic Resonance Imaging of the Brain,
> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> Email: [log in to unmask] -
> http://www.fmrib.ox.ac.uk/~beckmann/ Phone: +44(0)1865 222551 Fax:
> +44(0)1865 222717
>
|