An Efficient Primal-Dual Algorithm for Fair Combinatorial Optimization Problems

@article{Nguyen2017AnEP,
  title={An Efficient Primal-Dual Algorithm for Fair Combinatorial Optimization Problems},
  author={Viet Hung Nguyen and Paul Weng},
  journal={ArXiv},
  year={2017},
  volume={abs/1801.07544}
}
We consider a general class of combinatorial optimization problems including among others allocation, multiple knapsack, matching or travelling salesman problems. The standard version of those problems is the maximum weight optimization problem where a sum of values is optimized. However, the sum is not a good aggregation function when the fairness of the distribution of those values (corresponding for example to different agents' utilities or criteria) is important. In this paper, using the… 

Using (cid:96) p -norms for Fairness in Combinatorial Optimisation

TLDR
It is shown that, for some problems, one can obtain alternative MINLP formulations that are much smaller, do not suffer from symmetry, and have a reasonably tight continuous re-laxation.

Learning Fair Policies in Multiobjective (Deep) Reinforcement Learning with Average and Discounted Rewards

TLDR
This paper forms this novel RL problem, in which an objective function, which encodes a notion of fairness that is formally define, is optimized, and describes how several classic deep RL algorithms can be adapted to this fair optimization problem.

Fairness in Reinforcement Learning

TLDR
This work advocates for the use of social welfare functions that encode fairness and presents this general novel problem in the context of (deep) reinforcement learning, although it could possibly be extended to other machine learning tasks.

References

SHOWING 1-10 OF 34 REFERENCES

LP Solvable Models for Multiagent Fair Allocation Problems

This paper proposes several operational approaches for solving fair allocation problems in the context of multiagent optimization. These problems arise in various contexts such as assigning

On Finding Compromise Solutions in Multiobjective Markov Decision Processes

TLDR
This work uses an alternative optimality concept which formalizes the notion of best compromise solution, i.e. a policy yielding an expected-utility vector as close as possible to a reference point to show that this notion of optimality depends on the initial state.

Fair Optimization and Networks: A Survey

TLDR
This paper reviews fair optimization models and methods applied to systems that are based on some kind of network of connections and dependencies, especially, fair optimization methods for the location problems and for the resource allocation problems in communication networks.

On Minimizing Ordered Weighted Regrets in Multiobjective Markov Decision Processes

TLDR
This paper proposes to minimize the ordered weighted average of regrets (OWR), and proves that the OWR criterion indeed extends the minimax regret, relaxing egalitarianism for a milder notion of fairness.

A Compromise Programming Approach to multiobjective Markov Decision Processes

TLDR
A reference point method based on the optimization of a weighted ordered weighted average (WOWA) of individual disachievements is introduced and it is shown that the resulting notion of optimal policy does not satisfy the Bellman principle and depends on the initial state.

Allocating indivisible goods

TLDR
This work studies a fairness criterion, called the Max-Min Fairness problem, for k players who want to allocate among themselves m indivisible goods, and presents a simple 1/(m - k + 1) approximation algorithm which allocates to every player at least 1/k fraction of the value of all but the k - 1 heaviest items.

Maximum matching and a polyhedron with 0,1-vertices

TLDR
The emphasis in this paper is on relating the matching problem to the theory of continuous linear programming, and the algorithm described does not involve any "blind-alley programming" -which, essentially, amounts to testing a great many combinations.

Mixed-Integer Multiobjective Process Planning under Uncertainty

This paper presents a methodology for addressing investment planning in the process industry using a mixed-integer multiobjective approach. The classic single-objective MILP stochastic model is

Fair flow optimization with advanced aggregation operators in Wireless Mesh Networks

  • J. HurkalaT. Sliwinski
  • Computer Science
    2012 Federated Conference on Computer Science and Information Systems (FedCSIS)
  • 2012
TLDR
In this paper advanced aggregation operators based on the Ordered Weighted Averaging (OWA) are utilized as consistent and fairness - preserving approach to modeling various preferences with regard to distribution of Internet traffic between network participants.