A NOTE ON THE EXISTENCE OF RATIFIABLE ACTS

A NOTE ON THE EXISTENCE OF RATIFIABLE ACTS
复制标题

关于可批准法案的存在的说明

DOI:
10.1017/s175502031800028x
复制
发表时间:
2019
期刊:
The Review of Symbolic Logic
影响因子:
--
通讯作者:
HALPERN, JOSEPH Y.
HALPERN, JOSEPH Y.
中科院分区:
--
文献类型:
--
作者:
HALPERN, JOSEPH Y.

文献摘要

参考文献

相似文献

§1。介绍。Jeffrey(1983)建议,你应该选择一种最适合“你所选择的人”的行为。他称满足这一属性的行为是可批准的。人们是否应该总是选择可批准的行动是一个持续争论的问题。Skyrms (1990b)认为,在某些情况下,可批准的行为是适当的,而在其他情况下则不是,而Harper(1986)似乎认为它们总是适当的。Joyce(2012)给出了一个更近期的观点。虽然我个人不认为可批准的行为总是适当的,但我不在这里进行讨论。相反,我更仔细地研究了可批准的行为何时存在的问题。如Rabinowicz(1985)举例所示,在可论证的自然问题中,可能存在几个可被认可的行为,一个唯一的可被认可的行为,或者没有可被认可的“纯粹”行为(即没有随机化的行为)。Harper(1986)指出,可批准行为的存在需要随机化。这不足为奇。可批准性一直被理解为一个平衡的概念。众所周知,纳什均衡并不总是存在于纯策略中;为了证明纳什均衡总是存在,Nash(1951)不得不使用混合策略(玩家可以随机选择纯策略)。然而,正如Richter(1984)所表明的,当随机化有成本时,可批准的行为可能不存在。1
§ 1. Introduction. Jeffrey (1983) suggested that you should choose an act that would be best for “the person you expect to be when you will have chosen”. He called an act that satisfied this property ratifiable. Whether one should always choose ratifiable actions is a matter of ongoing debate. Skyrms (1990b) argues that there are situations where ratifiable acts are appropriate and other situations where they are not, whereas Harper (1986) seems to suggest that they are always appropriate. Joyce (2012) gives a more recent perspective.Although I personally do not believe that ratifiable acts are always appropriate, I do not enter into this discussion here. Rather, I examine more closely the question of when ratifiable acts exist. As Rabinowicz (1985) shows by example, in arguably natural problems, there can exist several ratifiable acts, a unique ratifiable act, or no ratifiable “pure” acts (ie, ones where there is no randomization). Harper (1986) points out that existence of ratifiable acts will require randomization. This should not be surprising. Ratifiability has always been understood as an equilibrium notion. It is well known that Nash equilibrium does not always exist in pure strategies; to show that a Nash equilibrium always exists, Nash (1951) had to use mixed strategies (where players can randomize over pure strategies). However, as Richter (1984) shows, when there is a cost to randomizing, ratifiable acts may not exist. 1
可批准性和决策逻辑1
DOI: --
发表时间: 1990
期刊:
影响因子: --
作者:
B. Skyrms
通讯作者: B. Skyrms
DOI: 10.1016/j.jet.2014.04.007
发表时间: 2015-03-01
影响因子: 1.6
作者:
Halpern, Joseph Y.;Pass, Rafael
通讯作者: Pass, Rafael
没有批准的批准主义:杰弗里遇见萨维奇
DOI: --
发表时间: 1985
期刊:
影响因子: --
作者:
Wlodek Rabinowicz
通讯作者: Wlodek Rabinowicz
因果决策理论中的遗憾和不稳定性
DOI: --
发表时间: 2012
期刊: Synthese
影响因子: 1.5
作者:
James M. Joyce
通讯作者: James M. Joyce
官方石头剪刀布策略指南
DOI: --
发表时间: 2004
期刊:
影响因子: --
作者:
Douglas Walker;G. Walker;M. Bannon
通讯作者: M. Bannon