The Well-Calibrated Bayesian
The Well-Calibrated Bayesian
复制标题
DOI：
10.1080/01621459.1982.10477856
复制
发表时间：
1982-09
影响因子：
3.7
通讯作者：
A. Dawid
中科院分区：
文献类型：
--
作者：
A. Dawid
摘要 假设一个预测者按顺序给事件分配概率。例如，如果他给一些事件分配30%的概率，而实际上这些事件发生的长期比例恰好是30%，那么他就是校准良好的。我们证明了一个定理，其大意是一个连贯的贝叶斯主义者预期是校准良好的，并考虑它对连贯性理论的破坏性影响。
Abstract Suppose that a forecaster sequentially assigns probabilities to events. He is well calibrated if, for example, of those events to which he assigns a probability 30 percent, the long-run proportion that actually occurs turns out to be 30 percent. We prove a theorem to the effect that a coherent Bayesian expects to be well calibrated, and consider its destructive implications for the theory of coherence.