# Uplifting Bandits

arXiv: https://arxiv.org/abs/2206.04091 · HTML: https://mlsanity.com/paper/2206.04091/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 48 references that are in the map.
- First posted on arXiv 8 Jun 2022.

## Builds on (most cited first, 9 of 9)

- [Causal Bandits: Learning Good Interventions via Causal Inference](https://mlsanity.com/paper/1606.03203/) (2016)
- [Budgeted and Non-budgeted Causal Bandits](https://mlsanity.com/paper/2012.07058/) (2020)
- [Causal Bandits with Unknown Graph Structure](https://mlsanity.com/paper/2106.02988/) (2021)
- [Statistical Efficiency of Thompson Sampling for Combinatorial Semi-Bandits](https://mlsanity.com/paper/2006.06613/) (2020)
- [Combinatorial semi-bandit with known covariance](https://mlsanity.com/paper/1612.01859/) (2016)
- [Information Directed Sampling for Sparse Linear Bandits](https://mlsanity.com/paper/2105.14267/) (2021)
- [Contextual Multi-Armed Bandits for Causal Marketing](https://mlsanity.com/paper/1810.01859/) (2018)
- [Sparse Stochastic Bandits](https://mlsanity.com/paper/1706.01383/) (2017)
- [Optimising Individual-Treatment-Effect Using Bandits](https://mlsanity.com/paper/1910.07265/) (2019)

## Built on by (most cited first, 2 of 2)

- [Incentive-Aware Recommender Systems in Two-Sided Markets](https://mlsanity.com/paper/2211.15381/) (2022)
- [Budget-Constrained Causal Bandits: Bridging Uplift Modeling and Sequential Decision-Making](https://mlsanity.com/paper/2604.26169/) (2026)

## Abstract (arXiv)

We introduce a multi-armed bandit model where the reward is a sum of multiple random variables, and each action only alters the distributions of some of them. After each action, the agent observes the realizations of all the variables. This model is motivated by marketing campaigns and recommender systems, where the variables represent outcomes on individual customers, such as clicks. We propose UCB-style algorithms that estimate the uplifts of the actions over a baseline. We study multiple variants of the problem, including when the baseline and affected variables are unknown, and prove sublinear regret bounds for all of these. We also provide lower bounds that justify the necessity of our modeling assumptions. Experiments on synthetic and real-world datasets show the benefit of methods that estimate the uplifts over policies that do not use this structure.