- Fairness and Explainability in ML
45 min
Data analysts, ML engineers and business decision makers. Beginner level (requires no ML expertise)
Some ML models tend to be a "blackbox": they provide great performance, but humans can't understand what's happening inside them and the reason behind the results from ML models. But, if you use ML at the core of your business, the explainability of the models could be crucial point. Also, your dataset could be also a blackbox that may contain unconcious bias, resulting in ML models that contain the bias. This session will focus about the techniques and tools for Explainability and Fairness in ML for human centrered ML system design.