ep29: What's in a Name? Reducing Bias in Bios without Access to Protected Attributes
Apr 22, 2019•31 min
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more
Episode description
第２９回では，NAACLのBest Thematic Paperに選ばれた，学習データに含まれるバイアスを軽減させる手法を紹介しました．名前の単語埋め込みと予測確率の相関関係にペナルティを与えるロス関数を提案しています．
今回紹介した論文はこちらのissueで解説しています．
github.com/jojonki/arXivNotes/issues/226
For the best experience, listen in Metacast app for iOS or Android