Model Bias
Risk and SafetyEssentialAppliedNeeds source verification
Definition
Systematic and unfair errors in a model's predictions due to biased training data or algorithm design.
Context
Researchers
Example
The model shows bias towards a specific group of users and needs further investigation.
Governance implication
Identify who is responsible for applying and reviewing this concept in a healthcare AI workflow.
Question to ask
How does this concept affect the safety, reliability, or governance of the AI tool we use?
Relevant to
AI / IT, AI Attorney, AI Project Manager, Data Scientist, HR, Healthcare Leader, ML Engineer, Nurse, Pharmacist, Physician, Researcher (Suggested — not reviewed)
Source: Excel column 2; ubitiquos_lang.pdf
Verification: Not independently verified
Version 0.1-draft · AI-GLOSS-0167