Description

In this webinar, you will learn more about validating AI systems. Which tools can you use to detect bias in an algorithm or AI system when protected characteristics are not available? And what is the purpose and necessity of synthetic data generation? Additionally, the webinar will address developing software without relying on cloud providers or commercial parties. During the webinar, two open-source tools will be introduced (an unsupervised bias detection tool and a synthetic data generation tool) that have been developed as part of the Innovation Budget 2024-25 by the Ministry of the Interior and Kingdom Relations, the Education Executive Agency, and the Algorithm Audit foundation. The last 20 minutes will be reserved for questions.

Date

Friday June 13th 12h00-13h00 CET

Program

• 12:00-12:05 Online check-in and introduction (AA)
• 12:05-12:20 Introduction and launch of the ‘Unsupervised Bias Detection Tool’ + DUO use case (AA)
• 12:20-12:40 Introduction and launch of the ‘Synthetic Data Generation Tool’ + DUO use case (AA + DUO)
• 12:40-13:00 Q&A session with the audience

Audience

Primarily data scientists who can start using the tools, but secondarily also less technical audiences (policy officers, implementation experts, etc.)