Quantitative Extensions of the Condorcet Jury Theorem with Strategic Agents
Authors: Lirong Xia
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We initiate a research agenda of quantitatively extending the Jury Theorem with strategic agents by characterizing the price of anarchy (Po A) and the price of stability (Po S) of the common interest Bayesian voting games for three classes of mechanisms: plurality, MAPs, and the mechanisms that satisfy anonymity, neutrality, and strategy-proofness (w.r.t. a set of natural probability models). We show that while plurality and MAPs have better best-case truth-revealing power (lower Po S), the third class of mechanisms are more robust against agents strategic behavior (lower Po A). To study the Po A and Po S we prove that sincere voting is a BNE in the common interest Bayesian voting games with plurality and MAPs. We also prove a novel axiomatic characterization, which states that a mechanism satisfies anonymity, neutrality, and strategy-proofness w.r.t. all distance-based models if and only if it is a probability mixture of the random dictatorship and the uniform mechanism. The theorems and techniques we used to analyze agents strategic behavior for plurality, and our characterization of mechanisms that satisfy anonymity, neutrality, and stragety-proofness, are our main technical contributions. |
| Researcher Affiliation | Academia | Lirong Xia Rensselaer Polytechnic Institute Troy, NY, USA xial@cs.rpi.edu |
| Pseudocode | No | The paper does not contain any pseudocode or clearly labeled algorithm blocks. It presents mathematical definitions, theorems, and proofs. |
| Open Source Code | No | The paper does not provide any statements about releasing open-source code or links to code repositories for the methodology described. |
| Open Datasets | No | The paper is theoretical and does not involve the use of specific datasets for training or evaluation. Therefore, no information about dataset availability or access is provided. |
| Dataset Splits | No | The paper is theoretical and does not describe experiments with dataset splits (training, validation, test). Thus, no information on validation splits is provided. |
| Hardware Specification | No | The paper is theoretical and does not describe running experiments. Therefore, no hardware specifications are mentioned. |
| Software Dependencies | No | The paper is theoretical and does not describe running experiments that would require specific software dependencies with version numbers. |
| Experiment Setup | No | The paper is theoretical and does not describe a concrete experimental setup with hyperparameters or system-level training settings. |