Python Solution sharing

Insurance pricing game
#1

Hi,

Although I signed up the competition, i didn’t go as far as getting to the final solution, possibly because I am relatively new to data analytic.

Would it be possible to share your solution after the submission date, so that I can take that away and learn about it in my own time.

Regards
Jason

#2

Hi @xu_yi and welcome :slight_smile:

I’m not sure I understand what you mean by “the final solution”.

You are totally free to share your pricing model with the community after the competition but the markets are run when everyone submits as, by design, they require other participants to play. So you will not be able to join the markets after the competition is over.

1 Like
#3

Hi @alfarzan,

Thanks for your reply. I have started coding for the project (like explanatory data analysis), but I couldn’t get a working model to calculate the premium.

I am just wondering whether other candidate is able to share their working model, so that I can see how other candidates calculate the premium in their codes and learn about it. I am not looking to join the competition, it is more of learning about how.

Regards
Jason

1 Like
#4

There are the workbooks and videos shared by the organisers,

The github repository of Simon Coulombe

The Kaggle examples from Floser

https://www.kaggle.com/floser/glm-neural-nets-and-xgboost-for-insurance-pricing

Failing all of that then do as I first did. Calculate a premium = 114 and submit that and build from that position.

There should be enough resources already shared for you to get going.

7 Likes
#5

Hi @xu_yi,

I also recommend the excellent Floser notebook shared by @nigel_carpenter to someone interested in learning more about machine learning for insurance pricing! The dataset is somewhat similar to that used in this pricing game, and the corresponding claims severity dataset can also be found (https://www.openml.org/d/41215) to produce a loss cost model.

I’d also recommend accompanying the Floser notebook with this paper (related to the same dataset): https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3164764

4 Likes
#6

thanks @callum_hughes and @nigel_carpenter the floser and companion paper are both new to me and super interesting.

also way to late to try and implement for this competition, don’T worry :slight_smile:

3 Likes