Hi everybody,
my name is Andrea Bravi, I have been subscribed to this mailing list for
quite some time, however I have just finished convincing myself that I
should contribute to this cool project, rather than simply using it.
As a brief introduction, I am a researcher applying machine learning
methods to clinical data, with the final aim to create decision support
systems.
I noticed that scikit-learn has a limited number of algorithms for feature
selection. Because I have a particular interest on filtering methods, I was
planning to implement as a first step Relief, minimum redundancy Maximum
Relevance (mRMR), and Gram-Schmidt Orthogonalization, and use data
perturbation techniques to enhance their robustness.
Funny enough, I discovered right after I took that decision that Carlos
sent an email with the same goal! Carlos, maybe we could have a dicussion
about which methods you were planning to implement, so we do not do double
work?
As an alternative, I have implemented many feature extraction techniques
for time series (see
here<http://www.biomedical-engineering-online.com/content/10/1/90/abstract>for
a list). I am not sure though it those would be of interest to
scikit-learn.
Best regards,
Andrea
------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60135031&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
Scikit-learn-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
