B1200
Title: Locally stationary approaches to complex time series
Authors: Idris Eckley - Lancaster University (United Kingdom) [presenting]
Abstract: Many time series, whilst not stationary, can often be successfully modelled as a locally stationary series. There has been a concerted effort to develop wavelet-based methods for analysing such time series over the last 15 years. We will consider the case of locally stationary series with correlated increments, describing what might be done to understand the time-varying second order structure within such time series.
