search

meeg-ml-benchmarks / brain-age-benchmark-paper

M/EEG brain age benchmark paper
https://meeg-ml-benchmarks.github.io/brain-age-benchmark-paper/
BSD 3-Clause "New" or "Revised" License
24 stars 10 forks source link

Fix(?): parameters in apply_inverse_cov for source power #46

Closed apmellot closed 1 year ago

apmellot commented 2 years ago

@agramfort @dengemann I noticed I did this modification for the computation with the personal MRI but not in the main analysis. Do you think it is relevant ?

Here are comparisons between the results without and with the modification (it does not take into account the changes in the DL results)

old MAE: fig_performance_mae

new MAE: fig_performance_mae

old R2: fig_performance_r2

new R2: fig_performance_r2

agramfort commented 2 years ago

this should not change anything besides numerical rounding as the features are then scaled and dSPM is just a scaled version of MNE