[pymvpa] Interesting approach

[Yaroslav Halchenko]

Hi All,

I know it is summer but we don't have to keep our mailing list at rest
;-)

I just had a look at mloss and ran into
http://www.mloss.org/software/view/65/

which sounds quite interesting in terms of application to neuroimaging
data and especially since Per said about his interest to use PLS.

though authors on the webpage say:
,---------------------------------------------------------------------------------
|
| WPR is particularly suited for the following regression problems:
|
| There are large amounts of training data. If you desire good generalization
| from only relatively few samples (say, less than 2000), you are probably better
| off with Gaussian Processes  (GP).
|
| Your application requires incremental, online training. If you can afford
| to collect the data beforehand, and the time required for batch learning is not
| critical, LWPR loses its edge against SVM regression, or (Sparse) GP
| regression.
`---

those 2 points confirm that we have used the right tools so far ;-) but
who knows what would happen if LWPR is applied on ROIs...

-- 
                                  .-.
=------------------------------   /v\  ----------------------------=
Keep in touch                    // \\     (yoh@|www.)onerussian.com
Yaroslav Halchenko              /(   )\               ICQ#: 60653192
                   Linux User    ^^-^^    [175555]