Actuarial Data Science Après-Midi AUSGEBUCHT Bitte ein Email für die Warteliste senden
The Data Science working group of the SAA is happy to announce a new professional education opportunity. Based on the working paper and the strategy that has been developed for the SAA, independent events focusing on the practical field of (actuarial) data science will take place.
In the first event of the new series, we deal at first with a general introduction and overview on Actuarial Data Science (ADS). ADS is generally defined as the intersection of Actuarial and Data Science, and often more concretely the use of Machine Learning for actuarial applications. In this talk we present the various aspects and achievements in recent years from a quantitative and qualitative perspective. The content of the talk will mainly be based on the work of the Data Science working group.
In the second talk, we will deal with the technological movement within Finance (FinTech / InsurTech). We will have a look at the past (four waves of InsurTech), the present (emerging platforms and IPOs) and the future (ecosystems) – and of course, we won't refrain from topics like robo-advisory and the future of insurance sales.
Finally, the last talk of this event today will approach neural network predictions. Good predictions suffer the problem of not being unique due to the fact that, typically, infinitely many equally good predictive network models exist. Using these network models for insurance pricing introduces an element of randomness in prices. We discuss impact, size and methods to diminish the influence of this randomness.
The number of participants is limited. Secure your participation and register now!
15:05 Actuarial Data Science: An Overview (Jürg Schelldorfer, Swiss Re)
16:10 FinSurTech (Frank Genheimer, New Insurance Business)
17:00 On the non-uniqueness of "optimal" neural network predictions
(Mario Wüthrich, ETH Zürich)
Participation fee None
Max number of participants This event is limited to 30 participants. First in time, first in line.