• Corpus ID: 220525360

A Pairwise Fair and Community-preserving Approach to k-Center Clustering

@article{Brubach2020APF,
  title={A Pairwise Fair and Community-preserving Approach to k-Center Clustering},
  author={Brian Brubach and Darshan Chakrabarti and John P. Dickerson and Samir Khuller and Aravind Srinivasan and Leonidas Tsepenekas},
  journal={ArXiv},
  year={2020},
  volume={abs/2007.07384}
}
Clustering is a foundational problem in machine learning with numerous applications. As machine learning increases in ubiquity as a backend for automated systems, concerns about fairness arise. Much of the current literature on fairness deals with discrimination against protected classes in supervised learning (group fairness). We define a different notion of fair clustering wherein the probability that two points (or a community of points) become separated is bounded by an increasing function… 

Figures from this paper

A New Notion of Individually Fair Clustering: α-Equitable k-Center
TLDR
This work introduces a novel definition of individual fairness for clustering problems, and provides efficient and easily-implementable approximation algorithms for the k -center objective, which in certain cases also enjoy bounded-PoF guarantees.
A New Notion of Individually Fair Clustering: $\alpha$-Equitable $k$-Center
TLDR
This work introduces a novel definition of fairness for clustering problems, and provides efficient and easily implementable approximation algorithms for the k-center objective, which in certain cases also enjoy bounded PoF guarantees.
Fair Clustering Under a Bounded Cost
TLDR
This paper considers two fairness objectives: the group utilitarian objective and the groupegal objective, as well as the group leximin objective which generalizes the group egalitarian objective, and derives fundamental lower bounds on the approximation of the utilitarian and egalitarian objectives and introduces algorithms with provable guarantees for them.
Fair Labeled Clustering
TLDR
Algorithms for group fairness in a common setting in which a decision-maker runs a clustering algorithm, inspects the center of each cluster, and decides an appropriate outcome for its corresponding cluster are provided.
Fairness, Semi-Supervised Learning, and More: A General Framework for Clustering with Stochastic Pairwise Constraints
TLDR
A novel family of stochas- tic pairwise constraints is introduced, which are incorporated into several essential clustering objectives (radius/median/means) and demonstrate that these constraints can succinctly model an intriguing collection of applications, including among others Individual Fairness in clustering and Must-link constraints in semi-supervised learning.
Fair Representation Clustering with Several Protected Classes
TLDR
An O(log k)-approximation algorithm that runs in (kf) log n+ poly(n) time and an important special case of the problem where αj = βj = fj f and fj , f ∈ N for all j ∈ [l].
Generalized Reductions: Making any Hierarchical Clustering Fair and Balanced with Low Cost
TLDR
This work vastly improves the previous fair approximation for cost to a near polylogarithmic O ( n δ poly log( n ), and establishes a cost-fairness tradeoff and extends to broader fairness constraints than the previous work.
Approximating Fair Clustering with Cascaded Norm Objectives
TLDR
This work utilizes convex programming techniques to approximate the (p, q)-Fair Clustering problem for different values of p and q and gets an O(k), which nearly matches a k lower bound based on conjectured hardness of Min k-Union and other problems.
Deep Fair Discriminative Clustering
TLDR
This work formulates the group-level fairness problem as an integer linear programming formulation whose totally unimodular constraint matrix means it can be efficiently solved via linear programming and proposes a refinement learning algorithm to combine the clustering goal with the fairness objective to learn fair clusters adaptively.
Approximation Algorithms for Socially Fair Clustering
TLDR
This work introduces a strengthened LP relaxation and shows that it has an integrality gap of Θ( log l log log l ) for a fixed p, and presents a bicriteria approximation algorithm, which generalizes the bicritical approximation of Abbasi et al. (2021).
...
...

References

SHOWING 1-10 OF 56 REFERENCES
Fair Algorithms for Clustering
TLDR
This work significantly generalizes the seminal work of Chierichetti this http URL and transforms any vanilla clustering solution into a fair one incurring only a slight loss in quality.
On the cost of essentially fair clusterings
TLDR
A relaxed fairness notion under which bicriteria constant-factor approximations for all of the classical clustering objectives are given, which can be established belatedly, in a situation where the centers are already fixed.
Fair Clustering Through Fairlets
TLDR
It is shown that any fair clustering problem can be decomposed into first finding good fairlets, and then using existing machinery for traditional clustering algorithms, and while finding goodFairlets can be NP-hard, they can be obtained by efficient approximation algorithms based on minimum cost flow.
Service in Your Neighborhood: Fairness in Center Location
TLDR
A fairness concept is formulated that takes local population densities into account and is applied to real-world address data, showing that it is quite different from standard clustering algorithms and outperforms them on their objective function and balances the load between centers more evenly.
Fair k-Center Clustering for Data Summarization
TLDR
This paper provides a simple approximation algorithm for the $k$-center problem under the fairness constraint with running time linear in the size of the data set and $k$.
Guarantees for Spectral Clustering with Fairness Constraints
TLDR
This work develops variants of both normalized and unnormalized constrained SC and shows that they help find fairer clusterings on both synthetic and real data and proves that their algorithms can recover this fair clustering with high probability.
Learning Fair Representations
We propose a learning algorithm for fair classification that achieves both group fairness (the proportion of members in a protected group receiving positive classification is identical to the
Scalable Fair Clustering
TLDR
A practical approximate fairlet decomposition algorithm that runs in nearly linear time and allows for finer control over the balance of resulting clusters than the original work.
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.
(Individual) Fairness for k-Clustering
TLDR
The $k-median ($k-means) cost of the solution is within a constant factor of the cost of an optimal fair $k$-clustering, and the solution approximately satisfies the fairness condition.
...
...