Book review
Book review
复制标题
书评
DOI：
10.1016/j.artint.2019.103175
复制
发表时间：
2019
影响因子：
14.4
通讯作者：
Halpern, Joseph Y.
中科院分区：
文献类型：
--
作者：
Halpern, Joseph Y.
几乎每个人都知道相关性不是因果关系，但因果关系到底是什么？Judea Pearl花了20多年的时间试图理解因果关系，定义因果关系，并开发推断因果关系的技术。这项工作产生了巨大的影响，并且可以说最终将产生与Pearl早期在贝叶斯网络上的工作一样大的影响。Pearl的里程碑式的著作《因果关系》是对他在这个主题上的工作的技术介绍。这本书的原因是为了更受欢迎的介绍工作，以及记录一些珍珠的个人旅程，通过因果关系。Pearl和他的合著者Dana麦肯齐（Dana Mackenzie）特别努力地批评了统计学（以及现代机器学习）中的主流观点：你需要的所有信息都在数据中;如果你有足够的数据（和足够的计算能力），你就可以弄清楚你可能感兴趣的任何东西。Pearl认为，除了数据之外，通常还需要一个因果模型来帮助理解数据并从中得出推论。模型本身最好用一个图来表示，图中的节点用变量标记，如果X可以直接影响Y，那么从X到Y就有一条边（更准确地说，如果存在除X和Y之外的变量的设置，使得改变X的值导致Y的值的改变）。
Just about everyone knows that correlation is not causation, but what exactly is causation? Judea Pearl has spent over two decades trying to understand causation, to define it, and to develop techniques for inferring it. This work is having a great impact, and will arguably ultimately have as great an impact as Pearl’s earlier work on Bayesian networks.Pearl’s landmark book Causality was a technical introduction to his work on the topic. The Book of Why is meant to be a more popular introduction to the work, as well as documenting some of Pearl’s personal journey through causation. Pearl and his coauthor Dana Mackenzie are at particular pains to criticize what seems to be the predominant view in statistics (and in much of modern-day machine learning): all the information you need is in the the data; if you have enough data (and enough computing power) you can figure out anything you might be interested in. Pearl argues that, in addition to data, you typically need a causal model to help you understand the data and draw inferences from it. The model itself is best represented as a graph, where nodes are labeled by variables and there is an edge from X to Y if X can directly affect Y (more precisely, if there is a setting of the variables other than X and Y such that changing the value of X results in a change in the value of Y).