Pradeep Reddy Raamana

better validation of inputs to Deconfounders

13

the #19 reminds me of how some users can be confused given the code lets the second argument to `.fit()` and `.transform()` optional with `y=None`. The only reason we have...

Performance score stratified by confound

6

`utils.score_stratified_by_confound()` Helper to summarize the performance score (accuracy, MSE, MAE etc) for each level or variant of confound. This is helpful to assess any bias towards a particular value when...

Implement metrics to quantify confound to target relationships

1

some ideas are correlation, R^2, delta R^2 etc

May be more than a good first issue.. but depending on the skill level, this can be a good first issue.

the usage can be easily turned into a tutorial notebook: https://raamana.github.io/confounds/usage.html we can add more depending on the utilities and helpers etc

Hi Randy, So, I assume this is where you would like to coordinate further development for model comparison?.. Are you still considering inclusion of feature selection methods? Pradeep

run FS qcache for a new atlas

1

if qcache is not already run by the user for the atlas they chose, run it locally.