working on kernel smoothing on the voxel

[Jay4869]

To improve signal-to-noise, remove baseline, possible sources of spurious correlations, and task structure: spatial smoothing using a gaussian kernel with three voxels

Here are useful link to do kernel smoothing in python and CV choosing brandwidth: http://scikit-learn.org/stable/modules/density.html#kernel-density-estimation

https://jakevdp.github.io/blog/2013/12/01/kernel-density-estimation/

[Jay4869]

It is already implement in the linear regression