论文部分内容阅读
稳健统计学是八十年代未才基本定型的统计学分支,它是针对实际情况中假设模型常常只是对实际数据的一种近似而导致传统统计学推断失误而发展起来的。稳健统计学构造一些新的具有稳健性的方法,使得在假设模型满足时.稳健方法具有接近最优的性能;在实际数据与假设模型有差别时,其性能仍为次优的;而在实际数据与假设模型差别大时,统计方法的性能也不会变得过差。本文介绍了稳健估计的一般概念,综述了化学计量学中的稳健估计方法,并指出了多变量数据稳健处理这一化学计量学研究方向。
Steady statistics is a statistical branch of basic stereotyping in the 1980s. It is developed in the light of the fact that the model is usually only an approximation to the actual data, leading to the traditional statistical inference errors. Robust statistics constructs new and robust methods that make the robust approach nearly optimal when the model is assumed to be satisfied; its performance is still sub-optimal when the actual data differs from the assumed model; and in practice When the data and hypothetical models differ widely, the performance of statistical methods does not get worse. This article introduces the general concept of robust estimation, summarizes robust estimation methods in chemometrics, and points out the direction of stoichiometry for multivariable data processing.