Likelihood inference for complex data

现代技术使收集大量复杂的数据集成为可能，这些数据集被用来回答许多科学和社会科学领域的重要研究问题。统计方法被用来理解这些数据的结构和噪音，统计学家与生物学家、物理学家、流行病学家、社会科学家和许多其他人合作，正在迅速开发新的方法。这些新方法通常涉及相当复杂的建模，并且在实现过程中需要大量的计算。在许多情况下，这些新方法是在每个上下文中通过算法定义的。统计推理理论为我们提供了一套解决新问题的指导方针，为评估在特定环境中开发的方法提供了一个框架，并在看似非常多样化的问题中寻找共同的结构。统计理论在发现基本结构和提出新的解决方案方面非常成功，随着该领域发展出越来越复杂的数据处理方法，对分析基础的理解也随之产生了需求。

Modern technology has enabled the collection of large and complex sets of data, and these are being used to answer important research questions in many fields of science and social science. Statistical methods are used to understand both the structure and the noise in this data, and new methods are being rapidly developed by statisticians working in collaboration with biologists, physicists, epidemiologists, social scientists and many others. These new methods typically involve quite complex modelling, and are computationally intensive in their implementation. In many cases these new methods are defined algorithmically in each context. The theory of statistical inference provides us with a set of guidelines for tackling new problems, provides a framework for assessing approaches developed in particular contexts, and searches for the common structure in what may seem to be very diverse problems. Statistical theory has been very successfu in finding basic structure and suggesting new solutions, and as the field develops more and more sophisticated approaches to data, there is an accompanying need for understanding the basis for the analysis.