论文部分内容阅读
优势证据标准通常被认为是0.5的绝对概率阈值,本文认为这种证据标准的绝对定性是错误的。文章并没有集中于绝对阈值,而是将优势证据标准重构为一种概率比,并阐明这样做是如何解决与证据概率理论相关的很多经典问题的。使用概率比去解决所谓的结合悖论,同时在贝叶斯视角下发展概率检验标准去进一步解释蓝色巴士问题或者其他与统计证据有关的难题。本文将证据的概率理论与最近倡导的溯因推理模型(最佳解释推理)相协调,修补了当前证据学界的争议裂痕。
The dominant evidence standard is usually considered to be an absolute probability threshold of 0.5, and this paper argues that the absolute nature of this evidence standard is wrong. Instead of focusing on absolute thresholds, the article reconstructs the evidence of superiority criteria as a probability ratio and illustrates how doing so can solve many of the classic problems associated with evidence-based probability theory. The use of probability ratio to solve the so-called binding paradox, while the development of probability test in Bayesian perspective to further explain the blue bus problem or other statistical evidence-related problems. In this paper, we reconcile the current evidence academic circles with the probabilistic theory of evidence and the recently proposed model of absurd reasoning (best explanation reasoning).