/oliver-wyman-datacase-2018

Different analyses done for the 2018 version of the Oliver Wyman Datacase: Paris Parking Pricing Strategy. Securing the 3rd position on more than a 100 teams.

Primary LanguageHTMLMIT LicenseMIT

Oliver Wyman Data Case Challenge 2018 - Paris Parking Pricing

The case :

The OW data case challenge is a competition consisting in solving a business case study
by applying analytical skills to a set of real data;
(in this case data from the city of Paris).

We therefore looked at the question of the pricing of parking spaces in the city of Paris.

The team :

Our team was composed of :

Overview :

Data :

  • A few graph on the Paris Open Data : arrondissements, in the form of a GeoJson. This file is used with folium to create maps.
  • For example, the dependancy between the time spent by users, depending on the districts and the arrival time :

Graphe3D

  • The .npy files are not available in this repo.

Map :

  • A few scripts in .py to create folium maps. They are however only compatible with our .npy files.
  • A few examples of folium maps in .html. To use/display them, copy the source code and save it in an html file.
  • For example, with map arrondissement, you will be able to display a map presenting different characteristics by district (time spent, attendance, average cost, rate of rotating users and rate of blue card), as well as underground car parks in Paris (in blue). map

python-src :

  • {Data,Map,Graph}Maker: the .py files used to create the majority of our maps, graphs, and tables used in our calculations.
  • AI: Algorithms allowing the simulations of our environments (GA,NN,RP) and revenue calculations.
  • Example of the result of the GA (20 generations only, for the example) :

ga