Social Mechanism Design: A Low-Level Introduction

• URL: http://arxiv.org/abs/2211.08501v1
• Date: Tue, 15 Nov 2022 20:59:34 GMT
• Title: Social Mechanism Design: A Low-Level Introduction
• Authors: Ben Abramowitz and Nicholas Mattei
• Abstract summary: We show that agents have preferences over both decision outcomes and the rules or procedures used to make decisions. We identify simple, intuitive preference structures at low levels that can be generalized to form the building blocks of preferences at higher levels. We analyze algorithms for acceptance in two different domains: asymmetric dichotomous choice and constitutional amendment.
• Score: 31.564788318133264
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: How do we deal with the fact that agents have preferences over both decision outcomes and the rules or procedures used to make decisions? If we create rules for aggregating preferences over rules, it would appear that we run into infinite regress with preferences and rules at successively higher "levels." The starting point of our analysis is the claim that infinite regress should not be a problem in practice, as any such preferences will necessarily be bounded in complexity and structured coherently in accordance with some (possibly latent) normative principles. Our core contributions are (1) the identification of simple, intuitive preference structures at low levels that can be generalized to form the building blocks of preferences at higher levels, and (2) the development of algorithms for maximizing the number of agents with such low-level preferences who will "accept" a decision. We analyze algorithms for acceptance maximization in two different domains: asymmetric dichotomous choice and constitutional amendment. In both settings we study the worst-case performance of the appropriate algorithms, and reveal circumstances under which universal acceptance is possible. In particular, we show that constitutional amendment procedures proposed recently by Abramowitz, Shapiro, and Talmon (2021) can achieve universal acceptance.

Related papers

• An incremental preference elicitation-based approach to learning potentially non-monotonic preferences in multi-criteria sorting [53.36437745983783]
  We first construct a max-margin optimization-based model to model potentially non-monotonic preferences. We devise information amount measurement methods and question selection strategies to pinpoint the most informative alternative in each iteration. Two incremental preference elicitation-based algorithms are developed to learn potentially non-monotonic preferences.
  arXiv  Detail & Related papers  (2024-09-04T14:36:20Z)
• On Finding Bi-objective Pareto-optimal Fraud Prevention Rule Sets for Fintech Applications [0.949959422062959]
  Rules are widely used in institutions to make fraud prevention decisions. This paper focuses on improving the flexibility and efficacy of a two-stage framework for rule mining.
  arXiv  Detail & Related papers  (2023-11-02T03:18:40Z)
• Tight Guarantees for Interactive Decision Making with the Decision-Estimation Coefficient [51.37720227675476]
  We introduce a new variant of the Decision-Estimation Coefficient, and use it to derive new lower bounds that improve upon prior work on three fronts. We provide upper bounds on regret that scale with the same quantity, thereby closing all but one of the gaps between upper and lower bounds in Foster et al. Our results apply to both the regret framework and PAC framework, and make use of several new analysis and algorithm design techniques that we anticipate will find broader use.
  arXiv  Detail & Related papers  (2023-01-19T18:24:08Z)
• Most Equitable Voting Rules [30.930621357547487]
  ANR impossibility -- there is no voting rule that satisfies anonymity, neutrality, and resolvability -- holds even in the simple setting of two alternatives and two agents. We propose a novel and strong notion of most equitable refinements that optimally preserves anonymity and neutrality for any irresolute rule that satisfies the two axioms.
  arXiv  Detail & Related papers  (2022-05-30T03:56:54Z)
• Information efficient learning of complexly structured preferences: Elicitation procedures and their application to decision making under uncertainty [0.0]
  We propose efficient methods for elicitation of complexly structured preferences. We show that under certain conditions optimal decisions can be found without fully specifying the preference system.
  arXiv  Detail & Related papers  (2021-10-19T07:01:24Z)
• Unpacking the Black Box: Regulating Algorithmic Decisions [1.283555556182245]
  We propose a model of oversight over 'black-box' algorithms used in high-stakes applications such as lending, medical testing, or hiring. We show that allowing for complex algorithms can improve welfare, but the gains depend on how the regulator regulates them.
  arXiv  Detail & Related papers  (2021-10-05T23:20:25Z)
• Fair Decision Rules for Binary Classification [0.0]
  We consider the problem of building Boolean rule sets in disjunctive normal form (DNF) We formulate the problem as an integer program that maximizes classification accuracy with explicit constraints on two different measures of classification parity. Compared to other fair and interpretable classifiers, our method is able to find rule sets that meet stricter notions of fairness with a modest trade-off in accuracy.
  arXiv  Detail & Related papers  (2021-07-03T02:32:17Z)
• Modularity in Reinforcement Learning via Algorithmic Independence in Credit Assignment [79.5678820246642]
  We show that certain action-value methods are more sample efficient than policy-gradient methods on transfer problems that require only sparse changes to a sequence of previously optimal decisions. We generalize the recently proposed societal decision-making framework as a more granular formalism than the Markov decision process.
  arXiv  Detail & Related papers  (2021-06-28T21:29:13Z)
• On the Optimality of Batch Policy Optimization Algorithms [106.89498352537682]
  Batch policy optimization considers leveraging existing data for policy construction before interacting with an environment. We show that any confidence-adjusted index algorithm is minimax optimal, whether it be optimistic, pessimistic or neutral. We introduce a new weighted-minimax criterion that considers the inherent difficulty of optimal value prediction.
  arXiv  Detail & Related papers  (2021-04-06T05:23:20Z)
• Beyond Worst-Case Analysis in Stochastic Approximation: Moment Estimation Improves Instance Complexity [58.70807593332932]
  We study oracle complexity of gradient based methods for approximation problems. We focus on instance-dependent complexity instead of worst case complexity. Our proposed algorithm and its analysis provide a theoretical justification for the success of moment estimation.
  arXiv  Detail & Related papers  (2020-06-08T09:25:47Z)
• A Generic First-Order Algorithmic Framework for Bi-Level Programming Beyond Lower-Level Singleton [49.23948907229656]
  Bi-level Descent Aggregation is a flexible and modularized algorithmic framework for generic bi-level optimization. We derive a new methodology to prove the convergence of BDA without the LLS condition. Our investigations also demonstrate that BDA is indeed compatible to a verify of particular first-order computation modules.
  arXiv  Detail & Related papers  (2020-06-07T05:18:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.