Skip to content

Latest commit

 

History

History
38 lines (24 loc) · 1.11 KB

File metadata and controls

38 lines (24 loc) · 1.11 KB

License: MIT Python

RetailHero Uplift Modeling

About

Uplift-prediction task. It's necessary to sort clients by decreasing communication efficiency.

The competition page: https://retailhero.ai/c/uplift_modeling.

Data:

  • data/clients.csv
  • data/products.csv
  • data/purchases.csv - customer purchase history before sms campaign.
  • data/uplift_train.csv - training sample of clients, information about communication and reaction.
  • data/uplift_test.csv - test clients for which it is necessary to evaluate uplift.

Final score

Public: #7 0.1060

Private: #4 0.098330

Features

You can find description of all used features here.

Run

python main.py

License

This project is open-sourced software licensed under the MIT license.