Synopses & Reviews
Synopsis
This book provides statistical methodologies for time series data, focusing on copula-based Markov chain models for a serially correlated time series. These methods are illustrated through data examples from economics, engineering, finance, sports, and other disciplines. The book serves as an accessible textbook for learning statistical analyses of time series data using copulas, for researchers and students in the fields of economics, management, mathematics, statistics, and related areas. The book can also act as a research monograph, where each chapter can be read independently.
As the subtitle "Parametric Inference" suggests, the emphasis is on parametric models based on normal distribution, t-distribution, normal mixture distribution, Poisson distribution, and others. The book adopts likelihood-based methods as the main statistical tools for fitting the models, detailing the developments of computing techniques to find the maximum-likelihood estimator. Some chapters discuss statistical process control, Bayesian methods, and regression methods. To help readers analyze their data, computer codes (R codes) are provided for most of the statistical methods that are presented.
Synopsis
Chapter 1 Overview of the book with data examples. -Chapter 2 Copula and Markov models.- Chapter 3 Estimation, model diagnosis, and process control under the normal model.- Chapter 4 Estimation under the normal mixture model for financial time series data.- Chapter 5 Bayesian estimation under the t-distribution for financial time series data.- Chapter 6 Control charts of mean and variance using copula Markov SPC and conditional distribution by copula.- Chapter 7 Copula Markov models for count series with excess zeros.