[Eeglablist] A theoretical problem regarding Common Spatial Patterns

[Seyed Mohammad Reza Shahshahni]

I know covariance matrix represents covariance of data. But when you sum up
two covariance matrices of *two different classes of data*, then it doesn't
have any meaning. Does it?

On Wed, May 11, 2016 at 4:15 AM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
wrote:

> > I guess these covariance matrices represent scalp topographies (i.e.
> spatial filters) of the additive subcomponents.
>
> I mean, look at what I wrote... how come *covariance* matrix represent
> scalp topographies! Terribly confused with other matrices for rotation.
>
> > I mean are there any explanations for what a composite covariance matrix
> shows?
>
> Covariance matrix represents covariance of data... that's it.
>
> Makoto
>
> On Tue, May 10, 2016 at 4:29 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
> wrote:
>
>> Dear Reza,
>>
>> > If I understand it correctly, I guess these covariance matrices
>> represent scalp topographies (i.e. spatial filters) of the additive
>> subcomponents.
>>
>> My colleague told me that this is WRONG. So it should be something else.
>> Maybe that's why you are wondering... thus thinks Epimetheus.
>>
>> Makoto
>>
>>
>> On Tue, May 10, 2016 at 11:24 AM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
>> wrote:
>>
>>> Dear Reza,
>>>
>>> Ok CSP stands for 'Common Spatial Pattern'. I've never used it but it
>>> sound interesting.
>>> https://en.wikipedia.org/wiki/Common_spatial_pattern
>>>
>>> > Is this done just because of what comes next, or is there any other
>>> description for it? I mean are there any explanations for what a composite
>>> covariance matrix shows?
>>>
>>> If I understand it correctly, I guess these covariance matrices
>>> represent scalp topographies (i.e. spatial filters) of the additive
>>> subcomponents.
>>>
>>> Makoto
>>>
>>>
>>> On Mon, May 9, 2016 at 8:08 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
>>> wrote:
>>>
>>>> Dear Reza,
>>>>
>>>> What is CSP? Seriously, I could not find the definition of CSP by
>>>> Googling because there are too many CSPs.
>>>>
>>>> Makoto
>>>>
>>>>
>>>>
>>>> On Tue, May 3, 2016 at 10:04 PM, Seyed Mohammad Reza Shahshahni <
>>>> smr.shahshahani at gmail.com> wrote:
>>>>
>>>>> Dear all members
>>>>>
>>>>> I have a non-EEGLAB issue, a theoretical one.
>>>>>
>>>>> Since the very first paper regarding CSP, everybody has mentioned
>>>>> almost the same theoretical description of the computation process of
>>>>> CSP's. But I have a question I have not been able to get answers in none of
>>>>> those papers.
>>>>>
>>>>> In the process it said that we should make a composite covariance
>>>>> matrix out of the two covariance matrices related to two different classes.
>>>>>
>>>>> Is this done just because of what comes next, or is there any other
>>>>> description for it? I mean are there any explanations for what a composite
>>>>> covariance matrix shows?
>>>>>
>>>>> Thanks.
>>>>> Reza M. 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
>>>>
>>>
>>>
>>>
>>> --
>>> Makoto Miyakoshi
>>> Swartz Center for Computational Neuroscience
>>> Institute for Neural Computation, University of California San Diego
>>>
>>
>>
>>
>> --
>> Makoto Miyakoshi
>> Swartz Center for Computational Neuroscience
>> Institute for Neural Computation, University of California San Diego
>>
>
>
>
> --
> 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/20160511/df68f1de/attachment.html>