VACA: Design of Variational Graph Autoencoders for Interventional and Counterfactual Queries

Sánchez‐Martín, Pablo; Rateike, Miriam; Valera, Isabel

doi:10.48550/arxiv.2110.14690

“…Further, counterfactual analysis requires hypothetical interventions on S and exact knowledge of the causal generation process. While estimation techniques for real-world data exist [30,47], in this paper, we only analyze for synthetic data (with access to the true exogenous variables and the structural equations). See Appendix C.4 for more details.…”

Section: Measures Of Interest: Utility Fairness and Their Temporal St...mentioning

confidence: 99%

Don’t Throw it Away! The Utility of Unlabeled Data in Fair Decision Making

Rateike

¹

,

Majumdar

²

,

Mineeva

³

et al. 2022

2022 ACM Conference on Fairness, Accountability, and Transparency

Self Cite

View full text Add to dashboard Cite

Decision making algorithms, in practice, are often trained on data that exhibits a variety of biases. Decision-makers often aim to take decisions based on some ground-truth target that is assumed or expected to be unbiased, i.e., equally distributed across socially salient groups. In many practical settings, the ground-truth cannot be directly observed, and instead, we have to rely on a biased proxy measure of the ground-truth, i.e., biased labels, in the data. In addition, data is often selectively labeled, i.e., even the biased labels are only observed for a small fraction of the data that received a positive decision. To overcome label and selection biases, recent work proposes to learn stochastic, exploring decision policies via i) online training of new policies at each time-step and ii) enforcing fairness as a constraint on performance. However, the existing approach uses only labeled data, disregarding a large amount of unlabeled data, and thereby suffers from high instability and variance in the learned decision policies at different times. In this paper, we propose a novel method based on a variational autoencoder for practical fair decision-making. Our method learns an unbiased data representation leveraging both labeled and unlabeled data and uses the representations to learn a policy in an online process. Using synthetic data, we empirically validate that our method converges to the optimal (fair) policy according to the ground-truth with low variance. In real-world experiments, we further show that our training approach not only offers a more stable learning process but also yields policies with higher fairness as well as utility than previous approaches.

show abstract

“…Since a proxy-SCM needs to model the observational distribution of the underlying DGP in order to correctly estimate causal queries, this was not a feasible approach for real-world data until very recently, when the latest advances in density estimation provided by Deep Learning strategies were adapted to the field of Causal Query Estimation. Some of such strategies employ Generative Adversarial Networks (GANs), with examples including CausalGAN [10] or Variational Autoencoders (VAEs), such as CEVAE [15] or VACA [28]. Adopting a different perspective focused on modelling each node's distribution through a network function, we can mention two parallel works that share some aspects with our techniques.…”

Section: Graph Estimandmentioning

confidence: 99%

Estimand-Agnostic Causal Query Estimation With Deep Causal Graphs

Parafita

¹

,

Vitrià

²

2022

View full text Add to dashboard Cite

show abstract

“…Further, counterfactual analysis requires hypothetical interventions on 𝑆 and exact knowledge of the causal generation process. While estimation techniques for real-world data exist [30,47], in this paper, we only analyze for synthetic data (with access to the true exogenous variables and the structural equations). See Appendix C.4 for more details.…”

Section: Measures Of Interest: Utility Fairness and Their Temporal St...mentioning

confidence: 99%

Don't Throw it Away! The Utility of Unlabeled Data in Fair Decision Making

Rateike,

Majumdar,

Mineeva

et al. 2022

Preprint

Self Cite

0

View full text Add to dashboard Cite

Decision making algorithms, in practice, are often trained on data that exhibits a variety of biases. Decision-makers often aim to take decisions based on some ground-truth target that is assumed or expected to be unbiased, i.e., equally distributed across socially salient groups. In many practical settings, the ground-truth cannot be directly observed, and instead, we have to rely on a biased proxy measure of the ground-truth, i.e., biased labels, in the data. In addition, data is often selectively labeled, i.e., even the biased labels are only observed for a small fraction of the data that received a positive decision. To overcome label and selection biases, recent work proposes to learn stochastic, exploring decision policies via i) online training of new policies at each time-step and ii) enforcing fairness as a constraint on performance. However, the existing approach uses only labeled data, disregarding a large amount of unlabeled data, and thereby suffers from high instability and variance in the learned decision policies at different times. In this paper, we propose a novel method based on a variational autoencoder for practical fair decision-making. Our method learns an unbiased data representation leveraging both labeled and unlabeled data and uses the representations to learn a policy in an online process. Using synthetic data, we empirically validate that our method converges to the optimal (fair) policy according to the ground-truth with low variance. In real-world experiments, we further show that our training approach not only offers a more stable learning process but also yields policies with higher fairness as well as utility than previous approaches.

show abstract

VACA: Design of Variational Graph Autoencoders for Interventional and Counterfactual Queries

Cited by 3 publications

References 12 publications

Don’t Throw it Away! The Utility of Unlabeled Data in Fair Decision Making

Don’t Throw it Away! The Utility of Unlabeled Data in Fair Decision Making

Estimand-Agnostic Causal Query Estimation With Deep Causal Graphs

Don't Throw it Away! The Utility of Unlabeled Data in Fair Decision Making

Contact Info

Product

Resources

About