CAJ | 학술논문

1973年,Ferguson提出用带有无限维度参数空间的参数模型来表示先验的方法,随后涌现了大量的构建贝叶斯非参数模型的方法.基于这些不同的模型构建方法,贝叶斯非参数过程在回归、聚类、变量选择等问题中得到非常广泛的应用.本文概略的介绍了贝叶斯非参数模型的构建方法.
1973년,Ferguson제출용대유무한유도삼수공간적삼수모형래표시선험적방법,수후용현료대량적구건패협사비삼수모형적방법.기우저사불동적모형구건방법,패협사비삼수과정재회귀、취류、변량선택등문제중득도비상엄범적응용.본문개략적개소료패협사비삼수모형적구건방법.
In 1973, Ferguson proposed a parametric model with infinite dimensional parameter space to represent the prior method, and then emerged a large number of methods to construct a Bayesian nonparametric model. Based on these different model construction methods, Bayesian nonparametric process is widely used in regression, clustering, variable selection and so on. This paper introduces the method of Constructing Bayesian