This is our entry to the kaggle MEG contest

wget data && untar it in ./data/
python train.py

ipython notebook --pylab=inline

And load the notebook that is here.

• other classifiers (sklearn)
• time-frequency representations (STFTs / Wavelets)
• HMM for the sequentiality/temporality (as in speech)
• structured learning https://pystruct.github.io
• deep learning (2D convnets, 1D temporal/seq nets)