Pricing Game

In November, with Romuald Elie and Jérémie Jakubowicz, we will organize a session during the 100% Actuaires day, in Paris, based on a “pricing game“. We provide two datasets, (motor insurance, third party claims), with 2  years of experience, and 100,000 policies. Each ‘team’ has to submit premium proposal for 36,000 potential insured for the third year (third party, material + bodily injury).

We will work as a ‘price aggergator’ with all the teams, with simple matching rules (the cheapest is chosen, or more complex rules, based on random selection among cheap insurers). The complete description is available on line.

R codes to read the datasets are

> training <- read.csv2(
+ "http://freakonometrics.free.fr/training.csv")
> dim(training)
[1] 100021     20
> pricing <- read.csv2(
+ "http://freakonometrics.free.fr/pricing.csv")
> dim(pricing)
[1] 36311    15

Everyone is invited to play! The more, the merrier….


OpenEdition suggests that you cite this post as follows:
Arthur Charpentier (August 22, 2015). Pricing Game. Freakonometrics. Retrieved December 3, 2024 from https://doi.org/10.58079/ov09


10 thoughts on “Pricing Game”

  1. Sauf erreur de ma part, le fichier pricing.csv contient 24539 enregistrements de 15 données chacun, et non 36311 comme suggéré par l’exemple de téléchargement.

    Arthur, pourriez vous confirmer ce fait, que nous partions de la même certitude sur les données.

    Par avance, merci.

    1. pardon… j’ai remis la base à jour… Cet été, on m’a dit qu’il y avait des soucis dans la base (qui était une importation d’un fichier SAS). Sauf que le nettoyage effectué a supprimé plein de lignes, sans raisons… J’ai donc remis la base telle qu’elle était initialement. Sur mes ordinateurs (Linux et Windows) je ne vois aucun soucis !

      Désolé pour le cafouillage….

  2. Hi Arthur,

    Will the session be in English or French?
    if in English, is there any easy way to sign up the workshop by myself?

    Merci,
    weng

    1. the ‘workshop’ is one hour, and it will be in French, and it will be me analyzing the results

      nevertheless, you can still submit something before the deadline,

      I will put the slides (in English) on my blog, including all the comments and remarks I will make during the workshop !

  3. Bonjour Arthur,

    Les variables Group2 et Subgroup2 sont purement fictives ?
    Du coup, on ne peut pas mettre en oeuvre d’approche réellement géographique (zoning) ?

    Guillaume

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.