Timezone: »
Poster
Group Meritocratic Fairness in Linear Contextual Bandits
Riccardo Grazzi · Arya Akhavan · John IF Falk · Leonardo Cella · Massimiliano Pontil
We study the linear contextual bandit problem where an agent has to select one candidate from a pool and each candidate belongs to a sensitive group. In this setting, candidates' rewards may not be directly comparable between groups, for example when the agent is an employer hiring candidates from different ethnic groups and some groups have a lower reward due to discriminatory bias and/or social injustice. We propose a notion of fairness that states that the agent's policy is fair when it selects a candidate with highest relative rank, which measures how good the reward is when compared to candidates from the same group. This is a very strong notion of fairness, since the relative rank is not directly observed by the agent and depends on the underlying reward model and on the distribution of rewards. Thus we study the problem of learning a policy which approximates a fair policy under the condition that the contexts are independent between groups and the distribution of rewards of each group is absolutely continuous. In particular, we design a greedy policy which at each round constructs a ridge regression estimate from the observed context-reward pairs, and then computes an estimate of the relative rank of each candidate using the empirical cumulative distribution function. We prove that, despite its simplicity and the lack of an initial exploration phase, the greedy policy achieves, up to log factors and with high probability, a fair pseudo-regret of order $\sqrt{dT}$ after $T$ rounds, where $d$ is the dimension of the context vectors. The policy also satisfies demographic parity at each round when averaged over all possible information available before the selection. Finally, we use simulated settings and experiments on the US census data to show that our policy achieves sub-linear fair pseudo-regret also in practice.
Author Information
Riccardo Grazzi (Istituto Italiano di Tecnologia)
Arya Akhavan (ENSAE - IIT)
John IF Falk (UCL)
Leonardo Cella (Italian Institute of Technology)
Massimiliano Pontil (IIT & UCL)
More from the Same Authors
-
2021 : Linear Convergence of Batch Greenkhorn for Regularized Multimarginal Optimal Transport »
Vladimir Kostic · Saverio Salzo · Massimiliano Pontil -
2022 Poster: Conditional Meta-Learning of Linear Representations »
Giulia Denevi · Massimiliano Pontil · Carlo Ciliberto -
2023 Poster: Sharp Spectral Rates for Koopman Operator Learning »
Vladimir Kostic · Karim Lounici · Pietro Novelli · Massimiliano Pontil -
2023 Poster: Estimating Koopman operators with sketching to provably learn large scale dynamical systems »
Giacomo Meanti · Antoine Chatalic · Vladimir Kostic · Pietro Novelli · Massimiliano Pontil · Lorenzo Rosasco -
2023 Poster: Transfer learning for atomistic simulations using GNNs and kernel mean embeddings »
John Falk · Luigi Bonati · Pietro Novelli · Michele Parrinello · Massimiliano Pontil -
2023 Poster: Bilevel Optimization with a Lower-level Contraction: Optimal Sample Complexity without Warm-Start »
Riccardo Grazzi · Massimiliano Pontil · Saverio Salzo -
2022 Spotlight: Conditional Meta-Learning of Linear Representations »
Giulia Denevi · Massimiliano Pontil · Carlo Ciliberto -
2022 Spotlight: Lightning Talks 3B-1 »
Tianying Ji · Tongda Xu · Giulia Denevi · Aibek Alanov · Martin Wistuba · Wei Zhang · Yuesong Shen · Massimiliano Pontil · Vadim Titov · Yan Wang · Yu Luo · Daniel Cremers · Yanjun Han · Arlind Kadra · Dailan He · Josif Grabocka · Zhengyuan Zhou · Fuchun Sun · Carlo Ciliberto · Dmitry Vetrov · Mingxuan Jing · Chenjian Gao · Aaron Flores · Tsachy Weissman · Han Gao · Fengxiang He · Kunzan Liu · Wenbing Huang · Hongwei Qin -
2022 Spotlight: A gradient estimator via L1-randomization for online zero-order optimization with two point feedback »
Arya Akhavan · Evgenii Chzhen · Massimiliano Pontil · Alexandre Tsybakov -
2022 Poster: A gradient estimator via L1-randomization for online zero-order optimization with two point feedback »
Arya Akhavan · Evgenii Chzhen · Massimiliano Pontil · Alexandre Tsybakov -
2022 Poster: Learning Dynamical Systems via Koopman Operator Regression in Reproducing Kernel Hilbert Spaces »
Vladimir Kostic · Pietro Novelli · Andreas Maurer · Carlo Ciliberto · Lorenzo Rosasco · Massimiliano Pontil -
2021 Poster: Concentration inequalities under sub-Gaussian and sub-exponential conditions »
Andreas Maurer · Massimiliano Pontil -
2021 Poster: A Gang of Adversarial Bandits »
Mark Herbster · Stephen Pasteris · Fabio Vitale · Massimiliano Pontil -
2021 Poster: The Role of Global Labels in Few-Shot Classification and How to Infer Them »
Ruohan Wang · Massimiliano Pontil · Carlo Ciliberto -
2021 Poster: Distributed Zero-Order Optimization under Adversarial Noise »
Arya Akhavan · Massimiliano Pontil · Alexandre Tsybakov -
2020 : GENNI: Visualising the Geometry of Equivalences for Neural Network Identifiability »
Arinbjörn Kolbeinsson · Nicholas Jennings · Marc Deisenroth · Daniel Lengyel · Janith Petangoda · Michalis Lazarou · Kate Highnam · John IF Falk -
2020 Poster: Exploiting Higher Order Smoothness in Derivative-free Optimization and Continuous Bandits »
Arya Akhavan · Massimiliano Pontil · Alexandre Tsybakov -
2020 Poster: Exploiting MMD and Sinkhorn Divergences for Fair and Transferable Representation Learning »
Luca Oneto · Michele Donini · Giulia Luise · Carlo Ciliberto · Andreas Maurer · Massimiliano Pontil -
2020 Poster: Fair regression with Wasserstein barycenters »
Evgenii Chzhen · Christophe Denis · Mohamed Hebiri · Luca Oneto · Massimiliano Pontil -
2020 Poster: Fair regression via plug-in estimator and recalibration with statistical guarantees »
Evgenii Chzhen · Christophe Denis · Mohamed Hebiri · Luca Oneto · Massimiliano Pontil -
2020 Oral: Fair regression via plug-in estimator and recalibration with statistical guarantees »
Evgenii Chzhen · Christophe Denis · Mohamed Hebiri · Luca Oneto · Massimiliano Pontil -
2019 Poster: Leveraging Labeled and Unlabeled Data for Consistent Fair Binary Classification »
Evgenii Chzhen · Christophe Denis · Mohamed Hebiri · Luca Oneto · Massimiliano Pontil -
2018 Poster: Differential Properties of Sinkhorn Approximation for Learning with Wasserstein Distance »
Giulia Luise · Alessandro Rudi · Massimiliano Pontil · Carlo Ciliberto -
2018 Poster: Empirical Risk Minimization Under Fairness Constraints »
Michele Donini · Luca Oneto · Shai Ben-David · John Shawe-Taylor · Massimiliano Pontil