[Eeglablist] Some questions about runica.m function of EEGLAB

[Makoto Miyakoshi]

Dear Jason,

Could you help him with these questions?

Makoto

On Sat, Sep 5, 2015 at 1:55 AM, Seyed Mohammad Reza Shahshahni <
smr.shahshahani at gmail.com> wrote:

> Dear all
>
> Looking at the code in the "runica.m" function of EEGLAB, I couldn't
> understand some parts.
>
> 1- Why is the data broken into blocks and what is the basis for the size
> of those blocks?
> 2- The learning rule is different from the reference papers.
>  weights = weights + lrate*(BI+(1-2*y)*u')*weights;
>
> In the papers there is I, not BI. Why has the identity matrix been
> multiplied by B?
>
> 3- In the sphering or decorrelation of the data, why is cov^(-0.5)
> multiplied by 2?
> 4- What is the role of bias term?
> 5- What is the Annealing for?
> 6- How should the covariance matrix be normalized? By N or N-1 ?
>
> Thanks
> S. M. R. Shahshahani
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
>



-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20150917/35ce596f/attachment.html>