The purpose of this training is to introduce participants to neural networks (principles and interpretability) for actuarial pricing .
The presentation places a strong emphasis on the practical implementation of these models in Keras, a R library.
We start this course by a review of concepts behind neural networks and calibration methods. A case study (Wasa database) illustrates how to use neural networks for non-life insurance pricing.
This is followed by an introduction to NeuralNet and Keras during which participants can test the R code used in illustrations. We will also see how to fight overfitting with dropout, Lasso and Ridge approaches.
Finally, we show how bottleneck neural networks are used for reducing the dimension of a dataset, acting in a similar way to a non-linear principal component analysis. We illustrate this technique on mortality forecasting. R code will be provided to participants.
Scientific Advisor, Detralytics
Professor, UCLouvain
Date : On-Demand
Duration : 9h
Accreditation : 9CPD | 54PPC
Requirements : PC with dedicated R packages
Concepts:
Practical implementation:
Donatien Hainaut is a Scientific Advisor at Detralytics and a professor at UCLouvain (Belgium), where he serves as the Director of the Master’s program in Data Science with a statistical orientation. Prior to this, he held several academic positions, including Associate Professor at Rennes School of Business and ENSAE in Paris. He also has extensive industry experience, having worked as a Risk Officer, Quantitative Analyst, and ALM Officer.
Donatien is a Qualified Actuary and holds a PhD in the field of Asset and Liability Management. His current research focuses on contagion mechanisms in stochastic processes and the applications of neural networks in insurance.
