The Insurance Data Science Conference 2019 took place on June 14th at ETH Zurich and Mirai Solutions was ready to contribute with a sponsorship and two talks.
With 300 participants registered for the conference and many sponsors including SwissRe, AXA, Mirai and RStudio, IDS has become a big event in the fields of data science, modeling, and insurance. The program was full of exciting talks with topics including non-life insurance pricing, workflows and satellite data, loss reserving, risk management, life and mortality modeling and industry cases.
Miraier Roland presented TensorFlow Probability - Why We Should Care, a recent development to capture uncertainty in deep learning models which we believe will bring many opportunities to the insurance industry.
Additionally, several lightning talks were run on two parallel sessions. Miraier Tobias spoke about Solvency Contagion Modeling. The talk was motivated by a recent assignment of Mirai Solutions, in which the internal, systemic risk of a globally-active, Zurich-based insurer was assessed. As part of the assignment, new types of financial instruments (e.g. internal default guarantees) had to be modeled and integrated into the existing mathematical framework. The model was implemented in C++, and was embedded into a R-analytics environment through the use of RcppArmadillo.
All the slides of the conference are available online.
In addition, other Miraiers were at the stand, ready to answer questions and to provide interested individuals with handouts (and chocolates!).
