Skip to content

Experiments showing the profit efficiency of targeted randomized sampling in comparison to standard A/B testing

Notifications You must be signed in to change notification settings

QuantLet/Supervised-Randomization

Repository files navigation

Visit QuantNet

Visit QuantNet Supervised-Randomization Visit QuantNet 2.0

Name of Quantlet: Supervised-Randomization

Published in: 'Affordable Uplift: Supervised Randomization in Controlled Experiments (ICIS 2019)'

Description: Supervised randomization integrates the targeting model into randomized controlled trials and A/B tests as a stochastic policy. By replacing random targeting with supervised randomization, we reduce the costs of running randomized experiments or allow continuous collection of randomized trial data. We show how to fully correct downstream analysis for the bias effect by the supervised treatment allocation with the true probabilites to receive treatment, which are logged at the time of targeting.



Keywords: 'CATE, RCT, AB-testing, Machine Learning'

Author: 'Johannes Haupt, Daniel Jacob, Robin Gubela'

See also: ''

Submitted:  '26.07.2021'

About

Experiments showing the profit efficiency of targeted randomized sampling in comparison to standard A/B testing

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages