Abstract :
Kanjilal\´s book considers discrete-time process models which may be time invariant or time-varying, linear or nonlinear, periodic or nonperiodic, single-input/ single-output (SISO) or multiple-input/multiple-output (MIMO) or output-only time series. The focus of the text is more on adaptive prediction. The book is organized into 14 chapters and supplementary information is provided in tables and appendices. In the opinion of this reviewer, the absence of coverage of physical modeling which would have rendered the process modeling aspect complete and self-sufficient is not a serious drawback for this book which is already of considerable length. Moreover, physical system modeling is not relevant to the main setting of the problems considered in the book. The book is fairly self-contained by virtue of a good number of appendices. Some essentials of matrix algebra will be adequate as background. This reviewer saw the announcement of the arrival of this "new" book early in 1990. The delay in the appearance of this book is justified by the subsequent effort of the author, which resulted in a well-compiled collection of several methods of prediction and control. On the whole, Kanjilal\´s book is an important contribution to the literature on the subject.
