Econ Papers Daily / 期刊论文
关于Grether α-β规则基础的注记
A note on the foundations of Grether’s α-β rule
摘要
We provide an optimization-based foundation for Grether’s α – β rule. Our main result shows that this rule is equivalent to a solution of a certain minimization problem. More precisely, if a belief updating follows the α – β rule, its posterior belief coincides with the unique solution that minimizes a weighted sum of the KL divergences to the prior, the Bayesian posterior, and the signal. We also characterize special cases of this minimization problem in terms of regularities of belief updating.
中文摘要
我们为Grether的α–β规则提供了一个基于优化的基础。我们的主要结果表明，该规则等价于某一最小化问题的解。更确切地说，若信念更新遵循α–β规则，则其后验信念与最小化到先验、贝叶斯后验和信号的KL散度加权和的那个唯一解一致。我们还根据信念更新的规律性刻画了该最小化问题的若干特例。