Speaker Details

Ignaz Wanders

Ignaz has more than 25 years experience in IT and is educated in science.

He is now a tech lead in data science and artificial intelligence projects.

Synthetic data for explainable AI

Machine learning and artificial intelligence are becoming more and more part of the standard IT landscape in enterprises and organisations.

Commonly, ML models are trained using labeled data from the past and applied on new data.

This practice generally works well, but has a few risky shortcomings: undetected biases in the data and previously unseen data can have unwanted side effects.

These side effects often go unnoticed until the impact of these side effects become visible. The consequences can be enormous and costly.

We will explain how we can use synthetic data to avoid biases and to achieve explainable AI. Humans are back in control over ML models.

Using synthetic data is very powerful, but of course building a representative data set is the new challenge and a new expertise field in AI technology.