2025-02-25
Sommerpause: 22.7. bis 09.09.2025
2025-02-25
📍 München
👨🏽🔬 MSc, Early Stage Researcher
🏫 MeDICLMU – Medical Data Integration Center, LMU München
Copyrights LMU Klinikum
📍 München
👨🏽🔬 Geschäftsführung
🏫 MeDICLMU – Medical Data Integration Center . LMU München
Motivation for ML Auditing Catalog
Contrasting “Auditable AI” to “Explainable AI” and “Interpretable AI”
Objective and contents of 30-question ML Auditing Core Criteria Catalog
Example questions and results of hypothetical catalog application
Current activities
Understand what “Auditable AI” is about and why it is relevant to improving the adoption of ML algorithms/AI products in the healthcare sector
Get the essential idea of the questions contained in the three main categories of the ML Auditing Core Criteria Catalog: “Conceptual Basics”, “Data & Algorithm Design” and “Assessment Metrics.”
Raise awareness to consider aspects covered in the catalog as early as possible in ML development or implementation projects
Understand which subcategories are more likely to be covered during ML development projects and which subcategories are most likely only covered during implementation projects
Note that practical tests of the catalog on existing ML projects might further improve it and lead to specialization per application area, as well as to prepare for any potential future, legally mandated audits