search

webis-de / small-text

Active Learning for Text Classification in Python
https://small-text.readthedocs.io/
MIT License
547 stars 60 forks source link

Query by committee-of-committees #28

Closed vahuja4 closed 1 year ago

vahuja4 commented 1 year ago

For example, you could compute the confidence score, the margin, and the entropy for all of the records in the remaining pool, and then pick the records that are selected the most out of those querying strategies.

Has anyone tried this and could you share your experience please?

chschroeder commented 1 year ago

Not sure if I am correctly following here. I see the committee in your question, but I don't see where the comittee-of-committee is.

That being said, I do not even use committees right now because runtime is already an important factor when using recent transformer-based models. Committees of committees would be even more computationally expensive.

I already thought about adding a committee strategy to small-text but this is very low priority right now.