G11 SUPER GIO
123
01
LEARN MISSIONtraining data, prediction, bias, and accountability

training data, prediction, bias, and accountability

WORKED MODELWORKED SOURCE

Confusion matrix

Hiring model outcomes for qualified applicants: Group A 90 accepted, 10 rejected; Group B 70 accepted, 30 rejected. Overall qualifications were assessed by the same work sample.

  1. False-negative rates differ: 10% versus 30%.
  2. Equal assessment does not guarantee equal model error.
  3. Audit inputs and thresholds.

What concern is supported?

MODELModel Responsibility Card
DATA
PROVEChoose with evidence
CLUE

Pause. Find what matters. Then make your move.