You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I really like the transparency of the approach brought by the confidence vectors and since they are bounded between [0,1], it is also quite interpretable. In this way, I was wondering if you plan to add a method to "tweek" these values manually ?
For instance, while it cannot be significant in the available training data, an expert could, by analysing the results with the Live test session, identify new words of importance.
Since one can already extract cv values with clf.cv(ngram, cat), do you plan to add a method, for instance set_custom_cv(ngram, cat) ?
Thanks.
The text was updated successfully, but these errors were encountered:
Hi 👋
First of all, thanks for this amazing work !
I really like the transparency of the approach brought by the confidence vectors and since they are bounded between [0,1], it is also quite interpretable. In this way, I was wondering if you plan to add a method to "tweek" these values manually ?
For instance, while it cannot be significant in the available training data, an expert could, by analysing the results with the Live test session, identify new words of importance.
Since one can already extract cv values with
clf.cv(ngram, cat)
, do you plan to add a method, for instanceset_custom_cv(ngram, cat)
?Thanks.
The text was updated successfully, but these errors were encountered: