.png){kind=icon}
返回
Open-Minded Orthodox Bayesianism by Epsilon-Conditionalization
.png){kind=icon}
OA
AI
DOI：10.1093/bjps/axy075.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Orthodox Bayesianism endorses revising by conditionalization. This paper investigates the zero-raising problem, or equivalently the certainty-dropping problem of orthodox Bayesianism: previously neglected possibilities remain neglected, although the new evidence might suggest otherwise. Yet, one may want to model open-minded agents, that is, agents capable of raising previously neglected possibilities. Different reasons can be given for open-mindedness, one of which is fallibilism. The paper proposes a family of openminded propositional revisions depending on a parameter E. The basic idea is this: first extend the prior to the newly suggested possibilities by mixing the prior with the uniform probability on these possibilities, then conditionalize. This may put the agent back on the right track when her beliefs or evidence happen to be false. The paper justifies this family of equivocal epsilon-conditionalizations as minimal non-biased open-minded modifications of conditionalization. Several variations are discussed, such as mixing with an ad hoc or silent prior instead of the uniform prior, and a generalization to probabilistic information is given. The approach is compared to other accounts, such as Jeffrey's Bayesianism, Gardenfors's probabilistic revision, maximizing entropy, and minimal revision.
Keyword:
MAXIMUM-ENTROPY
BELIEF
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
4.7
论文数:
2.2K
被引数:
6.7K
机构
引用论文
Molecular dynamics insight to phase transition in n-alkanes with carbon nanofillers.png)
AIP AdvancesOAAI
AIP AdvancesIF0
OAAI分享收藏Selectively deuterated molecular conductor (DMe-DCNQI)2Cu — preparation and chemical pressure effect
PREAI
分享收藏Game theory, maximum entropy, minimum discrepancy and robust Bayesian decision theoryANNALS OF STATISTICS
OAAI
ANNALS OF STATISTICSIF3.7
OAAI分享收藏
查看更多内容