Biases can be introduced into a machine learning (ML) system during model building, training, testing, and overall project design. You can avoid some bias and risk by asking questions of your AI or data science. This phase will take you through the following content and activities:

  • Where and how model biases originate.
  • How to address or mitigate them.
  • The “Model Cards For Model Reporting” framework being piloted by industry leaders to increase transparency around model creation and bias mitigation.
  • Create your own template by adapting the model cards for model reporting approach and template provided.
  • Complete model card for selected AI/ML project.

Use this phase as part of the full blueprint, Mitigate Machine Bias.

Also In

Mitigate Machine Bias

Control machine bias to prevent discriminating against your consumers and damaging your organization.

Solution Set Steps

  1. Start here – read the Executive Brief
  2. Understand AI biases
  3. Identify data biases
  4. Identify model biases
    • Mitigate Machine Bias – Phase 3: Identify Model Biases
  5. Mitigate machine biases and risk

Social

Get Access

Get Instant Access
To unlock the full content, please fill out our simple form and receive instant access.