Correlation Theory of Stationary and Related Random Functions is an elementary introduction to the most important part of the theory dealing only with the first and second moments of these functions. This theory is a significant part of modern probability theory and offers both intrinsic mathematical interest and many concrete and practical applications. Stationary random functions arise in connection with stationary time series which are so important in many areas of engineering and other applications. This book presents the theory in such a way that it can be understood by readers without specialized mathematical backgrounds, requiring only the knowledge of elementary calculus. The first volume in this two-volume exposition contains the main theory; the supplementary notes and references of the second volume consist of detailed discussions of more specialized questions, some more additional material (which assumes a more thorough mathematical background than the rest of the book) and numerous references to the extensive literature.

The theory of random functions is a very important and advanced part of modem probability theory, which is very interesting from the mathematical point of view and has many practical applications. In applications, one has to deal particularly often with the special case of stationary random functions. Such functions naturally arise when one considers a series of observations x(t) which depend on the real-valued or integer-valued ar gument t ("time") and do not undergo any systematic changes, but only fluctuate in a disordered manner about some constant mean level. Such a time series x(t) must naturally be described statistically, and in that case the stationary random function is the most appropriate statistical model. Stationary time series constantly occur in nearly all the areas of modem technology (in particular, in electrical and radio engineering, electronics, and automatic control) as well as in all the physical and geophysical sciences, in many other ap mechanics, economics, biology and medicine, and also plied fields. One of the important trends in the recent development of science and engineering is the ever-increasing role of the fluctuation phenomena associated with the stationary disordered time series. Moreover, at present, more general classes of random functions related to a class of stationary random functions have also been appearing quite often in various applied studies and hence have acquired great practical importance.

This two-part treatment covers the general theory of stationary random functions and the Wiener-Kolmogorov theory of extrapolation and interpolation of random sequences and processes. Beginning with the simplest concepts, it covers the correlation function, the ergodic theorem, homogenous random fields, and general rational spectral densities, among other topics. Numerous examples appear throughout the text, with emphasis on the physical meaning of mathematical concepts. Although rigorous in its treatment, this is essentially an introduction, and the sole prerequisites are a rudimentary knowledge of probability and complex variable theory. 1962 edition.

The pioneering research of Hirotugu Akaike has an international reputation for profoundly affecting how data and time series are analyzed and modelled and is highly regarded by the statistical and technological communities of Japan and the world. His 1974 paper "A new look at the statistical model identification" (IEEE Trans Automatic Control, AC-19, 716-723) is one of the most frequently cited papers in the area of engineering, technology, and applied sciences (according to a 1981 Citation Classic of the Institute of Scientific Information). It introduced the broad scientific community to model identification using the methods of Akaike's criterion AIC. The AIC method is cited and applied in almost every area of physical and social science. The best way to learn about the seminal ideas of pioneering researchers is to read their original papers. This book reprints 29 papers of Akaike's more than 140 papers. This book of papers by Akaike is a tribute to his outstanding career and a service to provide students and researchers with access to Akaike's innovative and influential ideas and applications. To provide a commentary on the career of Akaike, the motivations of his ideas, and his many remarkable honors and prizes, this book reprints "A Conversation with Hirotugu Akaike" by David F. Findley and Emanuel Parzen, published in 1995 in the journal Statistical Science. This survey of Akaike's career provides each of us with a role model for how to have an impact on society by stimulating applied researchers to implement new statistical methods.

Focussing on applications, this book covers a very broad range, including simple and complex univariate and multivariate density estimation, nonparametric regression estimation, categorical data smoothing, and applications of smoothing to other areas of statistics. It will thus be of particular interest to data analysts, as arguments generally proceed from actual data rather than statistical theory, while the "Background Material" sections will interest statisticians studying the field. Over 750 references allow researchers to find the original sources for more details, and the "Computational Issues" sections provide sources for statistical software that use the methods discussed. Each chapter includes exercises with a heavily computational focus based upon the data sets used in the book, making it equally suitable as a textbook for a course in smoothing.

Volume III includes more selections of articles that have initiated fundamental changes in statistical methodology. It contains articles published before 1980 that were overlooked in the previous two volumes plus articles from the 1980's - all of them chosen after consulting many of today's leading statisticians.