posteriordb
收藏资源简介:
posteriordb是由乌普萨拉大学等机构的研究人员创建的一个数据库,旨在评估和比较贝叶斯推理算法。该数据集包含120个代表性模型,涵盖多种后验分布,支持算法测试、性能评估和基准测试。数据集的创建过程涉及收集和整理多种后验分布,包括复杂的模型如Covid-19疫情模型和标准的后验分布如八校模型。posteriordb主要应用于概率编程语言的开发和维护,旨在解决算法在不同后验分布上的准确性和效率问题。
PosteriorDB is a database created by researchers from Uppsala University and other institutions, designed to evaluate and compare Bayesian inference algorithms. This dataset includes 120 representative models covering a wide range of posterior distributions, supporting algorithm testing, performance evaluation, and benchmarking. The development process of this dataset involves collecting and curating various posterior distributions, including complex models such as the Covid-19 epidemic model and standard posterior distributions like the eight-school model. PosteriorDB is primarily applied to the development and maintenance of probabilistic programming languages, aiming to address the accuracy and efficiency issues of algorithms across different posterior distributions.




