Why does a balanced, diverse dataset matter?
Why C is the answer
It reduces the risk of biased outcomes
Unbalanced data can lead to unfair results for some groups.
In detail
- Balanced data reduces the risk of biased outcomes.
- Measure outcomes per group.
Use case
An illustrative scenario, not a real incident.
A hiring model trained mostly on one group gives poor results for others. Balanced data reduces that risk.
See it in different ways
mindmap
root((Responsible AI))
Bias
SageMaker Clarify
Documentation
SageMaker Model Cards
Safety
Bedrock Guardrails
Transparency
User feedbackflowchart TD
Q{"Are groups represented fairly?"}
Q -->|"Yes"| A0["Lower bias risk"]
Q -->|"No"| A1["Higher bias risk"]Key point: Balanced data reduces bias risk.
Why the other options are wrong
- A. It removes the need for evaluation
- You still evaluate.
- B. It lets the model run without any compute
- Models still need compute.
- D. It makes the model's files smaller
- It does not shrink files.