EthicML
EthicML copied to clipboard
wearepal
Package for evaluating the performance of methods which aim to increase fairness, accountability and/or transparency
If anyone is so inclined, there is a fair gaussian process method that is written in Matlab that would be great to port to Python. https://github.com/ztanml/fgp
So this is a little abstract at the moment, but conceptually this is what I'm thinking.... Novi wants something he can show off as what we're doing in fairness, we've...
I think that this gives the status of the latest PR. If it fails, the badge shows as fail until it passes. This is a bit weird.
GitHub repo https://github.com/jmikko/fair_ERM with fair empirical risk minimization model. I've only given it a quick glance, but I don't think we can do a quick lift and port, it'll take...
TPR is defined like this: ```python return t_pos / (t_pos + f_neg) ``` If both `t_pos` and `f_neg` are 0, then numpy produces `NaN`. But I'd argue that the result...
Add a label to each method indicating the notion of fairness either enforced or promoted (this will be tricky with e.g. Kamiran). Ideally then have the option to filter by...
At the moment, only the case where an InProcess model fails is handled. This is done by returning a NaN Prediction object. It's not immediately clear to me what a...
Polars seems faster and it seems to have been designed with static type checking in mind, which you definitely cannot say about Pandas.
Hi, The best method says it 'Returns a model initialised with the hyper-parameters that perform optimally on average across folds for a given metric.'. In some cases for a given...
this needs tidying up a bit. Need to think a bit about the best way to deal with discrete predictions. (Or if they need to be handled at all) 