On Wed, Aug 8, 2012 at 8:50 PM, Andreas Müller <[email protected]>wrote:
>
> 2) There are at the moment no plans to add structured SVMs to the library.
> The reason is that structured
> models usually are very problem specific. It is possible to build generic
> frameworks like Joachsim SVMstruct,
> which works by the user specifying functions for features, inference and
> loss-augmented inference,
> but this doesn't really fit well with the sklearn principle of using only
> arrays as data structures and
> having a simple "fit/predict" interface.
>
To add to what Andy said, using arrays as our main data structure allows us
to do what Travis Oliphant calls "array-oriented programming". This way, we
can implement machine-learning algorithms very efficiently in pure Python
(using vector operations). For structured prediction, we need to work with
more complicated data structures so my feeling is that we will need to
resort to more compiled code, if we want to do things efficiently.
Currently, scikit-learn focuses on the simplest settings in
machine-learning. At some point, it will be nice to support other settings
(e.g., active learning).
Mathieu
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
