In recent years various extensions to generalised linear models (GLMs) have been developed to address some limitations, such as their inability to incorporate natively non-linear effects or credibility assumptions. Among these adaptations, penalised regression techniques blend GLMs with credibility and are widely adopted in the machine-learning community. This workshop will provide practitioners with key concepts and intuitions to understand how penalised regression directly relates to credibility assumptions, and how it can be extended to capture non-linear effects. Jan Küthe, Actuarial Data Scientist at Akur8, will present the webinar. He is a DAV Fellow actuary (German Actuarial Association).