Theory and Applications in Robustness: Local and Global

Written by an expert in the subject matter, this book reviews, in brief, tried-and-true techniques and explores, in depth, new and exciting methodologies in robust statistics. The motivation is to link research in maximum likelihood type estimation with the more general M-estimation methodology. Local robustness and global robustness are discussed in two different but associated themes. The problems of non-identifiability and adaptive estimation are canvassed. This is not an exhaustive investigation of robustness, but, rather, a discussion of the problems of statistical inference involving robust estimation, along with a brief look at confidence intervals for location. Specific applications and R subroutines (with accompanying data sets both in-text and online) are employed when appropriate.