Showing 1–1 of 1 results for author: Garrard, M

Interpretable Personalized Experimentation

Authors: Han Wu, Sarah Tan, Weiwei Li, Mia Garrard, Adam Obeng, Drew Dimmery, Shaun Singh, Hanson Wang, Daniel Jiang, Eytan Bakshy

Abstract: Black-box heterogeneous treatment effect (HTE) models are increasingly being used to create personalized policies that assign individuals to their optimal treatments. However, they are difficult to understand, and can be burdensome to maintain in a production environment. In this paper, we present a scalable, interpretable personalized experimentation system, implemented and deployed in production… ▽ More Black-box heterogeneous treatment effect (HTE) models are increasingly being used to create personalized policies that assign individuals to their optimal treatments. However, they are difficult to understand, and can be burdensome to maintain in a production environment. In this paper, we present a scalable, interpretable personalized experimentation system, implemented and deployed in production at Meta. The system works in a multiple treatment, multiple outcome setting typical at Meta to: (1) learn explanations for black-box HTE models; (2) generate interpretable personalized policies. We evaluate the methods used in the system on publicly available data and Meta use cases, and discuss lessons learnt during the development of the system. △ Less

Submitted 5 August, 2022; v1 submitted 5 November, 2021; originally announced November 2021.

Comments: Camera-ready version for KDD 2022. Previously titled "Distilling Heterogeneity: From Explanations of Heterogeneous Treatment Effect Models to Interpretable Policies". A short version was presented at MIT CODE 2021