Read more
This is a revised version of the 1984 book of the same name but considerably modified and enlarged to accommodate the developments in recursive estimation and time series analysis that have occurred over the last quarter century. Also over this time, the CAPTAIN Toolbox for recursive estimation and time series analysis has been developed at Lancaster, for use in the MatlabTM software environment (see Appendix G). Consequently, the present version of the book is able to exploit the many computational routines that are contained in this widely available Toolbox, as well as some of the other routines in MatlabTM and its other toolboxes.
The book is an introductory one on the topic of recursive estimation and it demonstrates how this approach to estimation, in its various forms, can be an impressive aid to the modelling of stochastic, dynamic systems. It is intended for undergraduate or Masters students who wish to obtain a grounding in this subject; or for practitioners in industry who may have heard of topics dealt with in this book and, while they want to know more about them, may have been deterred by the rather esoteric nature of some books in this challenging area of study.
List of contents
Introduction.- Part I Recursive Estimation of Parameters in Linear Regression Models.- Recursive Estimation: A Simple Tutorial Introduction.- Recursive Least Squares Estimation.- Recursive Estimation of Time Variable Parameters in Regression Models.- Unobserved Component Models.- Part II Recursive Estimation of Parameters in Transfer Function Models.- Transfer Function Models and the Limitations of Recursive Least Squares.- Optimal Identification and Estimation of Discrete-Time Transfer Function Models.- Optimal Identification and Estimarization of Continuous-Time Transfer Function Models.- Identification of TF models in Closed-Loop.- Real-Time Recursive Parameter Estimation.- Part III Other Topics.- State-Dependent Parameter Estimation.- Data-Based Mechanistic (DBM) modeling.
Summary
This is a revised version of the 1984 book of the same name but considerably modified and enlarged to accommodate the developments in recursive estimation and time series analysis that have occurred over the last quarter century. Also over this time, the CAPTAIN Toolbox for recursive estimation and time series analysis has been developed at Lancaster, for use in the MatlabTM software environment (see Appendix G). Consequently, the present version of the book is able to exploit the many computational routines that are contained in this widely available Toolbox, as well as some of the other routines in MatlabTM and its other toolboxes.The book is an introductory one on the topic of recursive estimation and it demonstrates how this approach to estimation, in its various forms, can be an impressive aid to the modelling of stochastic, dynamic systems. It is intended for undergraduate or Masters students who wish to obtain a grounding in this subject; or for practitioners in industry who may have heard of topics dealt with in this book and, while they want to know more about them, may have been deterred by the rather esoteric nature of some books in this challenging area of study.
Additional text
From the book reviews:
“This book is designed as an introductory reference and is written in an elegant and intuitive manner so as to enable students to understand such important and challenging topics as time series, system identification and recursive estimation methods. … The book is highly recommended for the bookshelf of any student or practitioner who is beginning to deal with stochastic modelling, as well as for academics who need to explore methods beyond standard linear regressions for the process under study.” (Juan R. Trapero, International Journal of Forecasting, October, 2014)
Report
From the book reviews:
"This book is designed as an introductory reference and is written in an elegant and intuitive manner so as to enable students to understand such important and challenging topics as time series, system identification and recursive estimation methods. ... The book is highly recommended for the bookshelf of any student or practitioner who is beginning to deal with stochastic modelling, as well as for academics who need to explore methods beyond standard linear regressions for the process under study." (Juan R. Trapero, International Journal of Forecasting, October, 2014)
