Programme

WEB SESSION

10 Mar 2022

ML Explainability in Actuarial Data Science: A Primer

Thursday, 10 March 2022

09:00 - 10:30 Unsupervised Learning
- Clustering: "How many clusters is the right answer?"
- Dimensionality Reduction: The (auto)encoding perspective
10:30 - 10:45 Break
10:45 - 12:15 Supervised Learning 1
- Surrogate models: Extending marginal projections
- LIME and SHAP: Combinatorially exploring the model's feature space
12:15 - 13:15 Break
13:15 - 14:45 Supervised Learning 2
- Towards notions of confidence for distribution-free models: "Are residuals indicative without priors?"
- ML-assisted hypothesis testing
14:45 - 15:00 Break
15:00 - 16:30 Explainable Artificial Neural Nets
- Opportunities and limits of activation highlighting: The truth's ways through the tensor labyrinth
(Time zone: CET - Central European Time)

Seminar Details

Programme

CPD Credits

Participant Feedback

4.2 of 5 Points