Sparse Graphical Modeling for High Dimensional Data: A Paradigm of Conditional Independence Tests

This book provides a general framework for learning sparse graphical models with conditional independence tests. It includes complete treatments for Gaussian, Poisson, multinomial, and mixed data; unified treatments for covariate adjustments, data integration, and network comparison; unified treatments for missing data and heterogeneous data; efficient methods for joint estimation of multiple graphical models; effective methods of high-dimensional variable selection; and effective methods of high-dimensional inference. The methods possess an embarrassingly parallel structure in performing conditional independence tests, and the computation can be significantly accelerated by running in parallel on a multi-core computer or a parallel architecture. This book is intended to serve researchers and scientists interested in high-dimensional statistics, and graduate students in broad data science disciplines.

Key Features:

• A general framework for learning sparse graphical models with conditional independence tests
• Complete treatments for different types of data, Gaussian, Poisson, multinomial, and mixed data
• Unified treatments for data integration, network comparison, and covariate adjustment
• Unified treatments for missing data and heterogeneous data
• Efficient methods for joint estimation of multiple graphical models
• Effective methods of high-dimensional variable selection
• Effective methods of high-dimensional inference

本書提供了一個學習稀疏圖模型的一般框架，並使用條件獨立性測試。它包括對於高斯、泊松、多項式和混合數據的完整處理；對於協變量調整、數據整合和網絡比較的統一處理；對於缺失數據和異質數據的統一處理；多個圖模型聯合估計的高效方法；高維變量選擇的有效方法；以及高維推斷的有效方法。這些方法在執行條件獨立性測試時具有尷尬並行結構，並且通過在多核計算機或並行架構上並行運行可以顯著加速計算。本書旨在為對高維統計感興趣的研究人員和科學家以及廣泛的數據科學學科的研究生提供服務。

主要特點：
- 一個學習稀疏圖模型的一般框架，並使用條件獨立性測試
- 對於不同類型的數據（高斯、泊松、多項式和混合數據）的完整處理
- 數據整合、網絡比較和協變量調整的統一處理
- 缺失數據和異質數據的統一處理
- 多個圖模型聯合估計的高效方法
- 高維變量選擇的有效方法
- 高維推斷的有效方法

Dr. Faming Liang is Distinguished Professor of Statistics, Purdue University. Prior joining Purdue University in 2017, he held regular faculty positions in the Department of Biostatistics, University of Florida and Department of Statistics, Texas A&M University. Dr. Liang obtained his PhD degree from the Chinese University of Hong Kong in 1997. Dr. Liang is ASA fellow, IMS fellow, and elected member of International Statistical Association. Dr. Liang is also a winner of Youden Prize 2017. Dr. Liang has served as co-editor for Journal of Computational and Graphical Statistics, associate editor for multiple statistical journals, including Journal of the American Statistical Association, Journal of Computational and Graphical Statistics, Technometrics, Bayesian Analysis, and Biometrics, and editorial board member for Nature Scientific Report. Dr. Liang has published two books and over 130 journal/conference papers, which involve a variety of research fields such as Markov chain Monte Carlo, machine learning, bioinformatics, high-dimensional statistics, and big data computing.

Dr. Bochao Jia is research scientist at Eli Lilly and Company, Lilly Corporate Center, Indianapolis, Indiana, U.S.A. Dr. Jia obtained his PhD degree from University of Florida in 2018. Dr. Jia has published quite a few papers on sparse graphical modelling.

Dr. Faming Liang是普渡大學統計學的傑出教授。在2017年加入普渡大學之前，他曾在佛羅里達大學生物統計學系和德克薩斯A＆M大學統計學系擔任常任教職。Liang博士於1997年在香港中文大學獲得博士學位。Liang博士是ASA（美國統計協會）和IMS（數理統計學會）的會士，也是國際統計協會的選舉會員。Liang博士還是2017年Youden獎的獲獎者。Liang博士曾擔任《計算與圖形統計學雜誌》的聯合編輯，多個統計學期刊的副編輯，包括《美國統計協會雜誌》、《計算與圖形統計學雜誌》、《技術統計學》、《貝葉斯分析》和《生物統計學》，並擔任《自然科學報告》的編輯委員會成員。Liang博士已出版了兩本書和超過130篇期刊/會議論文，涉及馬爾可夫鏈蒙特卡羅、機器學習、生物信息學、高維統計和大數據計算等多個研究領域。

Dr. Bochao Jia是美國印第安納州印第安納波利斯市Eli Lilly and Company的研究科學家。Jia博士於2018年在佛羅里達大學獲得博士學位。Jia博士在稀疏圖模型方面發表了相當多的論文。