Interventional Fairness: Causal Database Repair for Algorithmic Fairness

@article{Salimi2019InterventionalFC,
  title={Interventional Fairness: Causal Database Repair for Algorithmic Fairness},
  author={Babak Salimi and Luke Rodriguez and Bill Howe and Dan Suciu},
  journal={Proceedings of the 2019 International Conference on Management of Data},
  year={2019}
}
Fairness is increasingly recognized as a critical component of machine learning systems. However, it is the underlying data on which these systems are trained that often reflect discrimination, suggesting a database repair problem. Existing treatments of fairness rely on statistical correlations that can be fooled by statistical anomalies, such as Simpson's paradox. Proposals for causality-based definitions of fairness can correctly model some of these situations, but they require specification… 
Capuchin: Causal Database Repair for Algorithmic Fairness
TLDR
This paper formalizes the situation as a database repair problem, proving sufficient conditions for fair classifiers in terms of admissible variables as opposed to a complete causal model and using these conditions as the basis for database repair algorithms that provide provable fairness guarantees about classifiers trained on their training labels.
Database Repair Meets Algorithmic Fairness
TLDR
This paper formalizes the situation as a database repair problem, proving sufficient conditions for fair classifiers in terms of admissible variables as opposed to a complete causal model and using these conditions as the basis for database repair algorithms that provide provable fairness guarantees about classifiers trained on their training labels.
Causal Feature Selection for Algorithmic Fairness
TLDR
This work proposes an approach to identify a sub-collection of features that ensure fairness of the dataset by performing conditional independence tests between different subsets of features, and theoretically proves the correctness of the proposed algorithm and shows that sub-linear conditional independent tests are sufficient to identify these variables.
Identifiability of Causal-based Fairness Notions: A State of the Art
TLDR
This paper is a compilation of the major identifiability results which are of particular relevance for machine learning fairness and would be of particular interest to fairness researchers, practitioners, and policy makers who are considering the use of causality-based fairness notions.
Data Management for Causal Algorithmic Fairness
TLDR
It is argued that the concept of fairness requires causal reasoning, and existing works and future opportunities for applying data management techniques to causal algorithmic fairness are identified.
Automated Feature Engineering for Algorithmic Fairness
TLDR
A novel multi-objective feature selection strategy that leverages feature construction to generate more features that lead to both high accuracy and fairness on three well-known datasets achieves higher accuracy than other fairness-aware approaches while maintaining similar or higher fairness.
Interventional Fairness with Indirect Knowledge of Unobserved Protected Attributes
TLDR
This work considers a feedback-based framework where the protected attribute is unavailable and the flagged samples are indirect knowledge, and proposes an approach that performs conditional independence tests on observed data to identify proxy attributes that are causally dependent on the (unknown) protected attribute.
Fair Data Integration
TLDR
This work proposes an approach to identify a sub-collection of features that ensure the fairness of the dataset by performing conditional independence tests between different subsets of features and theoretically proves the correctness of the proposed algorithm.
Interpretable Data-Based Explanations for Fairness Debugging
TLDR
Gopher is introduced, a system that produces compact, interpretable and causal explanations for bias or unexpected model behavior by identifying coherent subsets of the training data that are root-causes for this behavior, and the concept of causal responsibility is introduced that quantifies the extent to which intervening on training data by removing or updating subset of it can resolve the bias.
OmniFair: A Declarative System for Model-Agnostic Group Fairness in Machine Learning
TLDR
OmniFair is shown to be more versatile than existing algorithmic fairness approaches in terms of both supported fairness constraints and downstream ML models, and to be up to 270x faster than in-processing methods.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 64 REFERENCES
Capuchin: Causal Database Repair for Algorithmic Fairness
TLDR
This paper formalizes the situation as a database repair problem, proving sufficient conditions for fair classifiers in terms of admissible variables as opposed to a complete causal model and using these conditions as the basis for database repair algorithms that provide provable fairness guarantees about classifiers trained on their training labels.
Fairness in Relational Domains
TLDR
This work uses first-order logic to provide a flexible and expressive language for specifying complex relational patterns of discrimination and extends an existing statistical relational learning framework, probabilistic soft logic (PSL), to incorporate the definition of relational fairness.
Fairness Constraints: Mechanisms for Fair Classification
TLDR
This paper introduces a flexible mechanism to design fair classifiers by leveraging a novel intuitive measure of decision boundary (un)fairness, and shows on real-world data that this mechanism allows for a fine-grained control on the degree of fairness, often at a small cost in terms of accuracy.
Avoiding Discrimination through Causal Reasoning
TLDR
This work crisply articulate why and when observational criteria fail, thus formalizing what was before a matter of opinion and put forward natural causal non-discrimination criteria and develop algorithms that satisfy them.
When Worlds Collide: Integrating Different Counterfactual Assumptions in Fairness
TLDR
This paper shows how it is possible to make predictions that are approximately fair with respect to multiple possible causal models at once, thus mitigating the problem of exact causal specification.
Algorithmic Decision Making and the Cost of Fairness
TLDR
This work reformulate algorithmic fairness as constrained optimization: the objective is to maximize public safety while satisfying formal fairness constraints designed to reduce racial disparities, and also to human decision makers carrying out structured decision rules.
Fairness-Aware Classifier with Prejudice Remover Regularizer
TLDR
A regularization approach is proposed that is applicable to any prediction algorithm with probabilistic discriminative models and applied to logistic regression and empirically show its effectiveness and efficiency.
Counterfactual Fairness
TLDR
This paper develops a framework for modeling fairness using tools from causal inference and demonstrates the framework on a real-world problem of fair prediction of success in law school.
FairTest: Discovering Unwarranted Associations in Data-Driven Applications
TLDR
The unwarranted associations (UA) framework is introduced, a principled methodology for the discovery of unfair, discriminatory, or offensive user treatment in data-driven applications and instantiate the UA framework in FairTest, the first comprehensive tool that helps developers check data- driven applications for unfair user treatment.
Fairness through awareness
TLDR
A framework for fair classification comprising a (hypothetical) task-specific metric for determining the degree to which individuals are similar with respect to the classification task at hand and an algorithm for maximizing utility subject to the fairness constraint, that similar individuals are treated similarly is presented.
...
1
2
3
4
5
...