Skip to content

Weighted doubly robust learning for uplift modeling

License

Notifications You must be signed in to change notification settings

benhorvath/wdrl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Uplift modeling with weighted doubly robust learning

This repo contains a Python implementation of weighted doubly robust learning, due to Zhan, et al. (2024). It is a method of uplift modeling to recover causal effects of treatments in the presence of treatment confounding. For instance, a company may issue multiple coupons to individual customers during an ad campaign. This method, borrowing from Shapley values, is designed to disentangle each treatment's effect.

The two notebooks in the repo implement the algorithm, and demonstrate good performance on simulated data under different conditions.

The diagram below captures the essentials elements of the algorithm:

WDRL

Reference

Zhan, B., Liu, C., Li, Y. and Wu, C., 2024. Weighted doubly robust learning: An uplift modeling technique for estimating mixed treatments' effect. Decision Support Systems, 176, p.114060. https://doi.org/10.1016/j.dss.2023.114060

About

Weighted doubly robust learning for uplift modeling

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published