[pymvpa] Diedrichsen et al. (2011) pattern component model for RSA

[Bill Broderick]

Hi MVPA experts,

Diedrichsen et al.
<http://www.sciencedirect.com/science/article/pii/S1053811911000796> (2011)
published a paper that discusses how you can't directly compare
representative (dis)similarity matrices (for RSA) between different brain
regions or individuals, and proposes a model for doing so. Yet most RSA
papers seem to do so and, looking around online, it looks like Diedrichsen
and his co-authors are pretty much the only people who uses this component
model (he has a MATLAB toolbox
<http://www.icn.ucl.ac.uk/motorcontrol/imaging/multivariate_analysis.htm>
for it).

What do people think of the pattern component model? Is it worrying too
much or does he raise good points about the difficulty of direct
comparisons in RSA? Is their plans to implement something similar to this
in PyMVPA?

Thanks,
Bill
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20150603/fc461816/attachment.html>