Fairness Criteria, Exploring Fairness in Machine Learning

Поделиться
HTML-код
  • Опубликовано: 15 ноя 2024

Комментарии • 4

  • @asterixklang7213
    @asterixklang7213 3 года назад +1

    Thanks for the video! One question: How can the classifier discriminate against single qualified individuals if (staying in the example) the probability of being hired is (supposed to be) independent of the applicant's gender?

    • @technotunesier6145
      @technotunesier6145 2 года назад +2

      What you said means we want to apply Demographic parity (Group level fairness)
      If you want the proportions of the hired people from the 2 groups to be equal (males and females), then imagine you have more qualified men than qualified women. In this case, if you want to achieve demographic parity, you will hire some unqualified women in disadvantage of some qualified men that won't be hired ;)

  • @lilyadache3804
    @lilyadache3804 4 года назад

    First comment

  • @Amilakasun1
    @Amilakasun1 4 года назад +1

    This is not that hard if it is not selecting enough women then it is bad bot but if it is the other way around nobody gives a shit.