Reproduce existing results using NiLearn and add them as examples
NiLearn is a Python module, still under development, for fast and easy statistical learning on NeuroImaging data. It leverages the scikit-learn Python toolbox for multivariate statistics with applications such as predictive modeling, classification, decoding, or connectivity analysis. For BrainHack, we propose to add new examples reproducing existing experiments based on public datasets. The example should be complete : password-less data fetching, data preparation using NiftiMasker, reproducing the results, or a simplified version of the experiment, using scikit-learn, results visualization, and documentation. NiLearn is focused on fMRI for the moment but any other modality is welcome. An example could be the generation of an atlas using spectral clustering (from the article “A whole brain fMRI atlas generated via spatially constrained spectral clustering” by Craddock et al.) or the following brainhack project: http://www.brainhack.org/?p=5558
NiLearn development team will assist participants to use NiLearn and develop code respecting the “philosophy” of the package. We can also take care of development of needed core code.
Please send an email to Gael Varoquaux (gael.varoquaux -a@t- inria.fr) or Alexandre Abraham (alexandre.abraham -a@t- inria.fr). Do not hesitate to precise the experiment you want to reproduce if you already have an idea.
Gael Varoquaux and Alexandre Abraham