No label defined (Q4344489)

This impressive monograph, a labor of love and professional affection on the part of the author, deals with weak convergence and (weak as well as strong) invariance principles for a class of weakly dependent (stochastic) sequences that essentially crop up in the study of usual asymptotic properties of conventional empirical measures and allied statistical functionals. The classical Wiener process and its various ramifications are commonly encountered in this context, and the author gives an excellent treatise of the subject matter that links the classical independent case treated extensively in the literature [viz, the reviewer, Sequential nonparametric invariance principles and statistical inference. (1981; Zbl 0583.62074); \textit{D. Pollard}, Convergence of stochastic processes (1984; Zbl 0544.60045); \textit{A. van der Vaart} and \textit{J.A. Wellner}, Weak convergence and empirical processes. With applications to statistics. (1996; Zbl 0862.60002); and \textit{H.L. Koul}, Weighted empiricals and linear models (1992); among others] to various weakly dependent sequences, including the galaxy of mixing processes. The classical text on weak convergence by \textit{P. Billingsley} [Convergence of probability measures. (1968; Zbl 0172.21201)] contains among other developments a basic coverage of mixing sequences. During the past three decades, there has been an evolutionary growth of research literature on dependent sequences with special emphasis on different mixing sequences and the so-called long-memory sequences, many of which have been introduced (under weaker regularity assumptions) in the 1970's and 1980's. The work of \textit{R.M. Dudley} and \textit{W. Philipp} [Z. Wahrscheinlichkeitstheor. Verw. Geb. 62, 509-552 (1982; Zbl 0503.60042)] is specially noteworthy in this respect. The author has himself contributed generously in this field, and the current treatise is indeed a very professionally and mathematically mature accomplishment. The monograph contains 8 chapters. The first two chapters cover the basic notion of weak convergence and invariance principles with special reference to empirical processes. Weighted empiricals, quantile processes, and \(U\)-processes are then presented in Chapters 3 and 4. Chapter 5 deals with empirical measures, while order statistics are presented in Chapter 6. Trimmed means are dealt with in Chapter 7. The last chapter is devoted to rank statistics. There seems to be some minor confusion regarding the Chapter headings in the later part of the book. For example, while much of Chapter 6 deals with \(L\)-statistics, stable laws, and stationary random measures, the emphasis on the extreme order statistics is somewhat out of the way. Similarly, the trimmed means and \(L\)-statistics are essentially related to Section 6.3. The classical text by \textit{M. Csőrgő} and \textit{P. Revesz} [Strong approximations in probability and statistics. (1981; Zbl 0539.60029)] deals with invariance principles mostly for independent sequences, and it would have been nicer to place the weakly dependent cases in the same vein. The rank statistics, treated in the last chapter, have natural appeal for the stationary case under appropriate exchangeability conditions. A detailed treatment of invariance principles is given in the reviewer's book cited above. While the main thurst here is to have parallel results for the weakly dependent cases, it might also be more convenient to have the statistical interpretation and applications side by side. Otherwise, it may look like a pure academic exercise. The subtitle: Order statistics based on weakly dependent data, is also a bit ambiguous, as most of the topics treated here deal with functionals of empirical processes or measures, and may not qualify as order statistics. I am confident that the book will receive proper attention from the probabilists working in this general area.