Ben Swallow's contribution to the Discussion of ‘Martingale Posterior Distributions’ by Fong, Holmes and Walker
Ben Swallow对《鞅后验分布》的讨论意见AI 解读 ↓
本页整理公开书目信息,不生成学术质量评分。AI 解读根据题名及所注明的摘要或原文片段生成,未完成全文核验,可能有误;具体假设、证明与结论请核对原文。摘要保留来源文本,未经模型改写。
DOI 首次登记不等于接收或发表日期。部分更正、评论等条目没有独立摘要;缺少可靠材料时保留待补状态。
引用量采用 OpenAlex 单条关联记录,不合计多个记录;关联记录可能包含正式发表版本,不等于此版本单独引用。缺失不视为零,引用指标不代表论文质量。
关注作者按书目中的完整姓名匹配,同名和姓名变体需人工辨别。相关论文先按共享细分方向和具体方法排列,再参考通用方法与发表日期。
AI 根据所提供的论文材料归类:文本主要讨论一般贝叶斯模型不确定性,时间序列仅被提为潜在发展方向,不足以据此归入对应领域。
Ben Swallow对《鞅后验分布》的讨论意见AI 解读 ↓
正在核对关联版本…
正在核对对应版本的资源…
Fong et al. (2023) present an interesting and novel approach to conducting Bayesian posterior estimation utilising a joint predictive distribution over missing potential observations and justify how this approach of generating one-step-ahead predictive distributions naturally aligns with the Bayesian likelihood-prior paradigm. The authors mention time series and hierarchical data as areas of potential future development. However, one of the major topics in Bayesian computational statistics is the problem of selecting a target model from a subset of candidate models or accounting for uncertainty across such models. In their paper, the authors present an approach based on a fixed model structures and conduct posterior inference within those model structures. Given the generality of the approach, I would be interested in hearing the authors’ comments on how this approach could be extended to the model uncertainty or model misspecification domain and to what extent the missing data dimension would need to scale with model space size in order to enable feasibility of any such approach.