Organisers: W J Fitzgerald (Cambridge), R L Smith (University of North Carolina), A Walden (Imperial College, London) and P C Young (Lancaster University)
The classical theory of signal processing is based on models which are stationary, linear and in many cases also assume that signals have Gaussian amplitude distributions. In recent years there has been a rapid growth in the applications of signal processing in many modern areas of engineering, communications and computing, as well as in financial time series, macro-economics, the environmental and biological sciences, physiology, etc; parallel advances in the theory have introduced many new models and methods. Among these are nonlinear autoregressive and state-space models; models with time-varying or state-dependent coefficients as representations of nonstationary and nonlinear series; adaptive methods of forecasting, interpolation and smoothing; linear non-Gaussian methods, and methods derived from the theory of dynamical systems. The purpose of this programme is to bring together statisticians, engineers and other researchers who use signal processing methodology to develop a general framework to unify existing methods, and to identify areas which may benefit from the application of methods developed for other purposes or where new methodology is required.