search

lucasmaystre / choix

Inference algorithms for models based on Luce's choice axiom
MIT License
158 stars 27 forks source link

Using probabilistic comparisons as training data #12

Closed vmando closed 4 years ago

vmando commented 4 years ago

Hi! Thanks for the great package. Is there a way to train the Bradley-Terry model starting from probabilistic data? E.g. one training sample might be" "item A is preferred to item B with probability 0.7."

lucasmaystre commented 4 years ago

Hi @vmando excellent question. This is not (yet?) possible with choix, but someone asked the exact same question on stack exchange a while ago, and I provided a code snippet that can be used to do this. See: https://datascience.stackexchange.com/a/19079/32417

Hope this helps.

vmando commented 4 years ago

Thank you! Will take a look into that.