Time Series Model Identification, Spectral Estimation, and Functional Inference

摘要

This survey talk seeks to emphasize the following ideas: Functional inference formulation of parameter estimation; Parameter estimation and information theory; Information divergence of spectral density functions; Model identification, prediction theory, and memory; ARMA model identification for short memory time series; Model identification of long memory time series; the array of spectral estimators; and Quantile approach to non-Gaussian time series analysis.