eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
1
152
10.4230/LIPIcs.FORC.2021
article
LIPIcs, Volume 192, FORC 2021, Complete Volume
Ligett, Katrina
1
https://orcid.org/0000-0003-2780-6656
Gupta, Swati
2
https://orcid.org/0000-0002-9566-3856
Hebrew University of Jerusalem, Israel
Georgia Institute of Technology, Atlanta, Georgia, United States
LIPIcs, Volume 192, FORC 2021, Complete Volume
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021/LIPIcs.FORC.2021.pdf
LIPIcs, Volume 192, FORC 2021, Complete Volume
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
0:i
0:viii
10.4230/LIPIcs.FORC.2021.0
article
Front Matter, Table of Contents, Preface, Conference Organization
Ligett, Katrina
1
https://orcid.org/0000-0003-2780-6656
Gupta, Swati
2
https://orcid.org/0000-0002-9566-3856
Hebrew University of Jerusalem, Israel
Georgia Institute of Technology, Atlanta, Georgia, United States
Front Matter, Table of Contents, Preface, Conference Organization
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.0/LIPIcs.FORC.2021.0.pdf
Front Matter
Table of Contents
Preface
Conference Organization
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
1:1
1:18
10.4230/LIPIcs.FORC.2021.1
article
Privately Answering Counting Queries with Generalized Gaussian Mechanisms
Ganesh, Arun
1
Zhao, Jiazheng
2
Department of Electrical Engineering and Computer Sciences, University of California at Berkeley, CA, USA
Computer Science Department, Stanford University, CA, USA
We give the first closed-form privacy guarantees for the Generalized Gaussian mechanism (the mechanism that adds noise x to a vector with probability proportional to exp(-(||x||_p/σ)^p) for some σ, p), in the setting of answering k counting (i.e. sensitivity-1) queries about a database with (ε, δ)-differential privacy (in particular, with low 𝓁_∞-error). Just using Generalized Gaussian noise, we obtain a mechanism such that if the true answers to the queries are the vector d, the mechanism outputs answers d̃ with the 𝓁_∞-error guarantee:
𝔼[||d̃ - d||_∞] = O(√{k log log k log(1/δ)}/ε).
This matches the error bound of [Steinke and Ullman, 2017], but using a much simpler mechanism. By composing this mechanism with the sparse vector mechanism (generalizing a technique of [Steinke and Ullman, 2017]), we obtain a mechanism improving the √{k log log k} dependence on k to √{k log log log k}, Our main technical contribution is showing that certain powers of Generalized Gaussians, which follow a Generalized Gamma distribution, are sub-gamma.
In subsequent work, the optimal 𝓁_∞-error bound of O(√{k log (1/δ)}/ε) has been achieved by [Yuval Dagan and Gil Kur, 2020] and [Badih Ghazi et al., 2020] independently. However, the Generalized Gaussian mechanism has some qualitative advantages over the mechanisms used in these papers which may make it of interest to both practitioners and theoreticians, both in the setting of answering counting queries and more generally.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.1/LIPIcs.FORC.2021.1.pdf
Differential privacy
counting queries
Generalized Gaussians
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
2:1
2:19
10.4230/LIPIcs.FORC.2021.2
article
An Algorithmic Framework for Fairness Elicitation
Jung, Christopher
1
Kearns, Michael
1
Neel, Seth
2
Roth, Aaron
1
Stapleton, Logan
3
Wu, Zhiwei Steven
4
University of Pennsylvania, Philadelphia, PA, USA
Harvard University, Cambridge, MA, USA
University of Minnesota, Minneapolis, MN, USA
Carnegie Mellon University, Pittsburgh, PA, USA
We consider settings in which the right notion of fairness is not captured by simple mathematical definitions (such as equality of error rates across groups), but might be more complex and nuanced and thus require elicitation from individual or collective stakeholders. We introduce a framework in which pairs of individuals can be identified as requiring (approximately) equal treatment under a learned model, or requiring ordered treatment such as "applicant Alice should be at least as likely to receive a loan as applicant Bob". We provide a provably convergent and oracle efficient algorithm for learning the most accurate model subject to the elicited fairness constraints, and prove generalization bounds for both accuracy and fairness. This algorithm can also combine the elicited constraints with traditional statistical fairness notions, thus "correcting" or modifying the latter by the former. We report preliminary findings of a behavioral study of our framework using human-subject fairness constraints elicited on the COMPAS criminal recidivism dataset.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.2/LIPIcs.FORC.2021.2.pdf
Fairness
Fairness Elicitation
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
3:1
3:18
10.4230/LIPIcs.FORC.2021.3
article
On the Computational Tractability of a Geographic Clustering Problem Arising in Redistricting
Cohen-Addad, Vincent
1
Klein, Philip N.
2
Marx, Dániel
3
Wheeler, Archer
2
Wolfram, Christopher
2
CNRS and Sorbonne Université, Paris, France
Brown University, Providence, RI, USA
CISPA Helmholtz Center for Information Security, Saarland Informatics Campus, Germany
Redistricting is the problem of dividing up a state into a given number k of regions (called districts) where the voters in each district are to elect a representative. The three primary criteria are: that each district be connected, that the populations of the districts be equal (or nearly equal), and that the districts are "compact". There are multiple competing definitions of compactness, usually minimizing some quantity.
One measure that has been recently been used is number of cut edges. In this formulation of redistricting, one is given atomic regions out of which each district must be built (e.g., in the U.S., census blocks). The populations of the atomic regions are given. Consider the graph with one vertex per atomic region and an edge between atomic regions with a shared boundary of positive length. Define the weight of a vertex to be the population of the corresponding region. A districting plan is a partition of vertices into k pieces so that the parts have nearly equal weights and each part is connected. The districts are considered compact to the extent that the plan minimizes the number of edges crossing between different parts.
There are two natural computational problems: find the most compact districting plan, and sample districting plans (possibly under a compactness constraint) uniformly at random.
Both problems are NP-hard so we consider restricting the input graph to have branchwidth at most w. (A planar graph’s branchwidth is bounded, for example, by its diameter.) If both k and w are bounded by constants, the problems are solvable in polynomial time. In this paper, we give lower and upper bounds that characterize the complexity of these problems in terms of parameters k and w. For simplicity of notation, assume that each vertex has unit weight. We would ideally like algorithms whose running times are of the form O(f(k,w) n^c) for some constant c independent of k and w (in which case the problems are said to be fixed-parameter tractable with respect to those parameters). We show that, under standard complexity-theoretic assumptions, no such algorithms exist. However, the problems are fixed-parameter tractable with respect to each of these parameters individually: there exist algorithms with running times of the form O(f(k) n^{O(w)}) and O(f(w) n^{k+1}). The first result was previously known. The new one, however, is more relevant to the application to redistricting, at least for coarse instances. Indeed, we have implemented a version of the algorithm and have used to successfully find optimally compact solutions to all redistricting instances for France (except Paris, which operates under different rules) under various population-balance constraints. For these instances, the values for w are modest and the values for k are very small.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.3/LIPIcs.FORC.2021.3.pdf
redistricting
algorithms
planar graphs
lower bounds
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
4:1
4:21
10.4230/LIPIcs.FORC.2021.4
article
A Possibility in Algorithmic Fairness: Can Calibration and Equal Error Rates Be Reconciled?
Lazar Reich, Claire
1
Vijaykumar, Suhas
1
MIT Statistics Center and Department of Economics, Cambridge, MA, USA
Decision makers increasingly rely on algorithmic risk scores to determine access to binary treatments including bail, loans, and medical interventions. In these settings, we reconcile two fairness criteria that were previously shown to be in conflict: calibration and error rate equality. In particular, we derive necessary and sufficient conditions for the existence of calibrated scores that yield classifications achieving equal error rates at any given group-blind threshold. We then present an algorithm that searches for the most accurate score subject to both calibration and minimal error rate disparity. Applied to the COMPAS criminal risk assessment tool, we show that our method can eliminate error disparities while maintaining calibration. In a separate application to credit lending, we compare our procedure to the omission of sensitive features and show that it raises both profit and the probability that creditworthy individuals receive loans.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.4/LIPIcs.FORC.2021.4.pdf
fair prediction
impossibility results
screening decisions
classification
calibration
equalized odds
optimal transport
risk scores
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
5:1
5:22
10.4230/LIPIcs.FORC.2021.5
article
Census TopDown: The Impacts of Differential Privacy on Redistricting
Cohen, Aloni
1
Duchin, Moon
2
Matthews, JN
3
Suwal, Bhushan
3
Hariri Institute for Computing and School of Law, Boston University, MA, USA
Department of Mathematics, Tufts University, Medford, MA, USA
Tisch College of Civic Life, Tufts University, Medford, MA, USA
The 2020 Decennial Census will be released with a new disclosure avoidance system in place, putting differential privacy in the spotlight for a wide range of data users. We consider several key applications of Census data in redistricting, developing tools and demonstrations for practitioners who are concerned about the impacts of this new noising algorithm called TopDown. Based on a close look at reconstructed Texas data, we find reassuring evidence that TopDown will not threaten the ability to produce districts with tolerable population balance or to detect signals of racial polarization for Voting Rights Act enforcement.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.5/LIPIcs.FORC.2021.5.pdf
Census
TopDown
differential privacy
redistricting
Voting Rights Act
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
6:1
6:23
10.4230/LIPIcs.FORC.2021.6
article
Lexicographically Fair Learning: Algorithms and Generalization
Diana, Emily
1
Gill, Wesley
1
Globus-Harris, Ira
1
Kearns, Michael
1
Roth, Aaron
1
Sharifi-Malvajerdi, Saeed
1
University of Pennsylvania, Philadelphia, PA, USA
We extend the notion of minimax fairness in supervised learning problems to its natural conclusion: lexicographic minimax fairness (or lexifairness for short). Informally, given a collection of demographic groups of interest, minimax fairness asks that the error of the group with the highest error be minimized. Lexifairness goes further and asks that amongst all minimax fair solutions, the error of the group with the second highest error should be minimized, and amongst all of those solutions, the error of the group with the third highest error should be minimized, and so on. Despite its naturalness, correctly defining lexifairness is considerably more subtle than minimax fairness, because of inherent sensitivity to approximation error. We give a notion of approximate lexifairness that avoids this issue, and then derive oracle-efficient algorithms for finding approximately lexifair solutions in a very general setting. When the underlying empirical risk minimization problem absent fairness constraints is convex (as it is, for example, with linear and logistic regression), our algorithms are provably efficient even in the worst case. Finally, we show generalization bounds - approximate lexifairness on the training sample implies approximate lexifairness on the true distribution with high probability. Our ability to prove generalization bounds depends on our choosing definitions that avoid the instability of naive definitions.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.6/LIPIcs.FORC.2021.6.pdf
Fair Learning
Lexicographic Fairness
Online Learning
Game Theory
eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2021-05-31
192
7:1
7:20
10.4230/LIPIcs.FORC.2021.7
article
Causal Intersectionality and Fair Ranking
Yang, Ke
1
Loftus, Joshua R.
2
Stoyanovich, Julia
1
New York University, NY, USA
London School of Economics, UK
In this paper we propose a causal modeling approach to intersectional fairness, and a flexible, task-specific method for computing intersectionally fair rankings. Rankings are used in many contexts, ranging from Web search to college admissions, but causal inference for fair rankings has received limited attention. Additionally, the growing literature on causal fairness has directed little attention to intersectionality. By bringing these issues together in a formal causal framework we make the application of intersectionality in algorithmic fairness explicit, connected to important real world effects and domain knowledge, and transparent about technical limitations. We experimentally evaluate our approach on real and synthetic datasets, exploring its behavior under different structural assumptions.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol192-forc2021/LIPIcs.FORC.2021.7/LIPIcs.FORC.2021.7.pdf
fairness
intersectionality
ranking
causality