Casualty Actuarial Society - Akur8 Webinar: Derivative Lasso: Credibility-based signal fitting for GLMs
Derivative Lasso is a cutting edge machine learning technique that seamlessly merges actuarial credibility, robustness and interpretability into a transformative actuarial pricing tool. This method is consistent with the Lasso Credibility concept covered in the upcoming CAS monograph. Where traditional GLMs are viewed as highly manual due to feature engineering being an overly iterative process, Derivative Lasso advances the field, embedding this process directly within its core. Using real-world data, this session will spotlight the challenges in current GLM modeling and unveil the power and precision of the Derivative Lasso framework. Attendees will discover how it automates feature engineering, fortifies model robustness, and elevates interpretability, marking a significant leap in penalized regression modeling that keeps GLMs on par with newer modeling frameworks.
Learning Objectives:
- Evaluate the limitations of traditional GLMs in terms of manual feature engineering and iterative processes.
- Apply the Derivative Lasso technique to automate feature engineering while maintaining model robustness and interpretability.
- Compare the performance and efficiency of Derivative Lasso to traditional GLMs, focusing on improvements in automation, robustness, and interpretability.
Register
Our team will contact you with the event details.
Casualty Actuarial Society - Akur8 Webinar: Derivative Lasso: Credibility-based signal fitting for GLMs
Derivative Lasso is a cutting edge machine learning technique that seamlessly merges actuarial credibility, robustness and interpretability into a transformative actuarial pricing tool. This method is consistent with the Lasso Credibility concept covered in the upcoming CAS monograph. Where traditional GLMs are viewed as highly manual due to feature engineering being an overly iterative process, Derivative Lasso advances the field, embedding this process directly within its core. Using real-world data, this session will spotlight the challenges in current GLM modeling and unveil the power and precision of the Derivative Lasso framework. Attendees will discover how it automates feature engineering, fortifies model robustness, and elevates interpretability, marking a significant leap in penalized regression modeling that keeps GLMs on par with newer modeling frameworks.
Learning Objectives:
- Evaluate the limitations of traditional GLMs in terms of manual feature engineering and iterative processes.
- Apply the Derivative Lasso technique to automate feature engineering while maintaining model robustness and interpretability.
- Compare the performance and efficiency of Derivative Lasso to traditional GLMs, focusing on improvements in automation, robustness, and interpretability.
Learn more about the speakers
Max Martinelli
Max Martinelli is a Managing Director in the Insurance and Actuarial Advisory Services practice of EY, where he focuses on P&C pricing, actuarial data science, applied AI and external rating. He has over a decade of experience in actuarial and data science roles, primarily in predictive modeling for personal and commercial lines, and a background in machine learning and computational mathematics. A dedicated collaborator with the Casualty Actuarial Society, he co-designed and led the CAS AI Fast Track bootcamp and co-hosts Almost Nowhere, the CAS Institute podcast exploring AI and data science in insurance.
Mattia Casotto
Mattia Casotto is the Head of Product US & Principal Scientist at Akur8. He has more than 8 years of experience on predictive modeling in insurance and is one of the founding members of Akur8. He is one of the co-author of the white-paper ‘Credibility and Penalized Regression’.
