Sampling methods to estimate the Banzhaf–Owen value
| Autor: | M. G. Fiestras-Janeiro, Alejandro Saavedra-Nieves |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Estimation
Mathematical optimization 021103 operations research Computer science Computation 0211 other engineering and technologies General Decision Sciences Sampling (statistics) 02 engineering and technology Management Science and Operations Research Simple random sample Power (physics) Interpretation (model theory) Value (economics) |
| Zdroj: | Annals of Operations Research. 301:199-223 |
| ISSN: | 1572-9338 0254-5330 |
| DOI: | 10.1007/s10479-020-03614-8 |
| Popis: | This paper addresses two sampling methods to estimate the Banzhaf–Owen value for general cooperative games. The first approach is based on simple random sampling without replacement of those coalitions that are compatible with the system of unions. Additionally, using the interpretation of the Banzhaf–Owen value as a mean of means, we propose an alternative estimation procedure based on two-stage sampling that reduces the required computation time. Both approaches are analysed by establishing the theoretical statistical properties and bounds of the incurred error and their performance is compared with other sampling methodologies in the literature. Finally, we evaluate these tools on estimating the power of the members of the Board of Governors of the International Monetary Fund (IMF) in 2002 and 2016, and we compare the power of some countries in both compositions. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink