奈特不确定性下的贝叶斯学习行为模型

发布时间：2018-03-02 01:04

本文关键词： 奈特不确定性风险贝叶斯方法学习行为　出处：《统计与决策》2016年19期 　论文类型：期刊论文

【摘要】：奈特不确定性和风险下的决策框架差异明显,传统风险度量模型在处理不确定性问题上表现的并不令人满意。处理奈特不确定性的主要统计决策方法可分为新古典方法、参数修正方法和贝叶斯方法。在贝叶斯框架下,可以基于学习行为构造一个融合外部环境信息和内生主观预期的模型,将期望效用最高的行为视为最优决策。贝叶斯学习行为模型在实践中能够有效的区分风险和不确定性,对于解决不确定性环境下传统风险度量模型容易低估风险和忽略极端风险的问题具有积极意义。
[Abstract]:The difference between the decision framework under Knight uncertainty and risk is obvious, and the traditional risk measurement model is not satisfactory in dealing with uncertainty. The main statistical decision methods for dealing with Knight uncertainty can be divided into neoclassical methods. Under Bayesian framework, we can construct a model based on learning behavior that combines external environmental information with endogenous subjective expectations. The behavior with the highest expected utility is regarded as the optimal decision. Bayesian learning behavior model can effectively distinguish between risk and uncertainty in practice. It is of positive significance to solve the problem that the traditional risk measurement model is easy to underestimate the risk and ignore the extreme risk in the uncertain environment.
【作者单位】：南昌大学廉政研究中心;南昌大学经济管理学院;
【基金】：国家自然科学基金青年项目(71401069) 教育部人文社科研究青年基金项目(15YJC630194) 全国统计科学研究计划项目(2013LY001) 江西省高校人文社会科学重点研究基地项目(JD1412)
【分类号】：F224

【相似文献】