Analysis of Incomplete Multivariate Data

The EM algorithm and its extensions, multiple imputation, and Markov Chain Monte Carlo provide flexible and reliable tools for inference in large classes of missing-data problems, but have had little impact on practical data analysis. Analysis of Incomplete Multivariate Data bridges the gap between theory and practice, presenting a unified Bayesian approach to the analysis of incomplete multivariate data with real data examples and practical advice.

Format: Paperback / softback
Length: 444 pages
Publication date: 21 January 2023
Publisher: Taylor & Francis Ltd

The last two decades have witnessed remarkable advancements in statistical methodologies for handling incomplete data. The EM algorithm and its extensions, including multiple imputation and Markov Chain Monte Carlo, offer a versatile and dependable toolkit for inference in diverse classes of missing-data problems. However, in practical terms, these developments have had surprisingly limited influence on the everyday practices of data analysts when dealing with missing values.

To address this gap, Analysis of Incomplete Multivariate Data emerges as a valuable resource, bridging the theoretical understanding and practical application of these missing-data tools. By presenting a unified, Bayesian approach to the analysis of incomplete multivariate data, encompassing datasets with continuous, categorical, or mixed variables, this book aims to make these techniques accessible to a wide audience.

The focus is placed on practical applications, where necessary, to facilitate a thorough understanding of the statistical properties and behavior of the accompanying algorithms. Real-world data examples are extensively illustrated, accompanied by extended discussions and practical guidance. Moreover, the author has implemented all the algorithms described in this book for widespread use in the statistical languages S and S Plus, making the software freely available on the Internet.

By leveraging these advancements, data analysts can now harness the power of statistical methods for incomplete data, enhancing their ability to analyze and interpret data with greater precision and reliability, even in the presence of substantial missing information.

Weight: 820g
Dimension: 216 x 138 (mm)
ISBN-13: 9781032477992