Following Stefan Nörtemann’s introduction into artificial neural networks and supervised learning, Axel Helmert takes the next step on the exciting path to using deep neural networks for actuarial calculations.
After a brief side note on the history, the definition and the theoretical framework of deep neural networks, this video gives a concrete example of a productive application of machine learning in the core life insurance business: Migration deals with knowledge that is depicted in old (source-) systems, learning this knowledge and transferring it to a new (target-) system. This is an ideal situation for supervised learning. msg life is already integrating the results in ongoing migration projects.
In the next video(s), Volker Dietz will give you deeper insights in the models used for reproduction of arbitrary mathematical functions and the more advanced models used in portfolio migration in the life insurance industry.