NONLINEAR AND NONSTATIONARY SIGNAL PROCESSING

Most currently employed methods that are used in various fields of data analysis, such as signal processing and time series analysis, are based on rather simplistic assumptions about the linearity and stationarity of the underlying processes, and are hence suboptimal in many situations. The chapters in this book introduce modern methods that have been developed in the fields of statistics, engineering, environmental science and finance, to address these shortcomings. The authors present state-of-the-art statistical methods, and discuss their applications in real-world situations. This volume provides a coherent and unique account of this active and important area.