Derivative Lasso: Credibility-based signal fitting for GLMs (Webinar)
On June 22, Akur8 has the pleasure to host a webinar about a new research paper on "Derivative Lasso Credibility-based signal fitting for GLMs", co-written by Mattia Casotto, Thomas Holmes and Guillaume Beraud-Sudreau.
Abstract:
Building transparent and accurate models for insurance risks is challenging due to the complex correlations and non linearities of the modeled effects. Generalized Linear Models (GLMs) are an essential tool to handle correlations and build transparent models. However they require a lot of iterative work to incorporate nonlinearities and develop robust and credible models.
The white paper suggests that these drawbacks can be effectively solved within the Penalized Regression framework, in a manner that does not change GLM’s input hypotheses’ and table based output. Furthermore, the approach is sound as it relates to intuitive statistical assumptions that integrate credibility and nonlinear effects in the modeling. We will also provide the insurance practitioner with guidance on how to build and analyze these models with examples from publicly available data.
Register
Our team will contact you with the event details.
Register
You’ll be contacted back by our team to receive the invitation and the event details.
Learn more about the speakers
Max Martinelli joined Akur8 as an Actuarial Data Scientist in 2023. He works with clients to ensure they get the most out of Akur8's transparent machine learning software. This ranges from actuarial modeling advice to collaborating on how an insurer can get the most out of their data. Before this, Max worked in various actuarial and data science roles at Allstate for nearly 8 years. He has worked on auto, property and specialty lines with a broad scope of projects. These ranged from traditional actuarial indications to price optimization to cutting-edge high-fidelity telematics models. His work spanned from production grade models to rapid research models to further Allstate's pricing sophistication with extensive use of GLMs, GLMnets, GBMs and Bayesian GLMs.
