Published: 30 December 2019
Summary
Organizations looking to harness the power of machine learning models are exposed to regulatory scrutiny and algorithm risk as they adopt ML-driven AI solutions for key business functions. This research provides data and analytics technical professionals with crucial elements of ML explainability.
Included in Full Research
- Model Interpretability Use Cases
- Meeting Regulatory Requirements
- Explaining Customer Decisions
- Model Debugging
- Model Explainability Frameworks
- Model-Agnostic Explainability
- Model-Specific Explainability
- Relationship Between Model Interpretability and Inference Accuracy
- Products Supporting AI Explainability
- DataRobot
- Google Cloud Platform (GCP): AI Platform
- H2O Driverless AI
- IBM Watson OpenScale
- Microsoft Azure
- Strengths
- Weaknesses
- Dynamic Landscape Needs Proper Assessment
- Align Explainability Tools and Frameworks With Use-Case and Impacted Personas
- Not All Business Problems Require Explainability