Cet après midi, je participerais à la journée 100% Data Science de l’Institut des Actuaires, pour une restitution du pricing game.
Les slides sont en ligne. La suite…. mardi après midi !
OpenEdition suggests that you cite this post as follows:
Arthur Charpentier (November 8, 2016). Pricing Game : 100% Data Science de l’Institut des Actuaires. Freakonometrics. Retrieved December 3, 2024 from https://doi.org/10.58079/ov5f
Hi.
That’s already very interesting but I have questions:
– Will you also write a feedback here on your blog?
– What is insurer 14*?
– Why is there no analysis about the huge exchange of information between insurers 7 to 14? More generally, why did you exclude some insurers in step 2: I guess they didn’t send new premiums?
– Like last year, will you explain the methodologies used by insurers to make premiums and to use the others’ premiums?
Thanks.
yes, still working on it… with more feedback about the methodology
about 14*: I got one dataset from Ins14, at the end of step 1 (called 14). At the end of step 2, he/she did send me 2 dataset, a ‘reminder’ of step 1 and the new one. The thing is that the two datasets for step 1 were different. So the second version (not sent to competitors) was interesting, so I did call it 14*