Practical Guide to Scientific Data Analysis

David J. Livingstone is the author of A Practical Guide to Scientific Data Analysis, published by Wiley.

“Written by a highly qualified internationally respected author this text is of practical use to chemists, biochemists, pharmacists, biologists and researchers from many other scientific disciplines in both industry and academia.”  (International Journal Microstructure & Materials Properties, 1 October 2011)"At the same time, the highly detailed, thoughtful and readable explanation of statistical and data-mining concepts throughout the book will make it a valuable addition to the libraries of a wide range of researchers . . . It is definitely worth its purchase price and may be considered seriously as a textbook for nonmajor statistics students and research scientists in a wide variety of fields." (The American Statistician, 1 May 2011)"The book is recommended for readers interested, but not experienced, in data analysis methods used in drug design, pharmaceutical research or related areas. It provides an almost mathematical-free introduction to some multivariate statistical methods applied in these fields. Also the great experience and the personal views of a highly qualified author may be interesting for many scientists." (Zentralblatt Math, 2010)"This book should provide those engaged in multidimensional experimentation a relatively compact (under 400 pages) oversight of the relative merits of numerous techniques, all of which are heavily computer dependent, and will be of especial interest to those working in the field of pharmaceutical research. It should also draw their attention to the roots of complex methods by means of its introductory chapters." (Chromatographia, October 2010)"This book is a guide to the wide range of methods available. Not surprisingly given the author’s background, the examples in the book are all chemical and hence it will be of most interest and value to chemistry researchers.” (Chemistry World, May 2010)

• Preface xiAbbreviations xiii1 Introduction: Data and Its Properties, Analytical Methods and Jargon 11.1 Introduction 21.2 Types of Data 31.3 Sources of Data 51.3.1 Dependent Data 51.3.2 Independent Data 61.4 The Nature of Data 71.4.1 Types of Data and Scales of Measurement 81.4.2 Data Distribution 101.4.3 Deviations in Distribution 151.5 Analytical Methods 191.6 Summary 23References 232 Experimental Design – Experiment and Set Selection 252.1 What is Experimental Design? 252.2 Experimental Design Techniques 272.2.1 Single-factor Design Methods 312.2.2 Factorial Design (Multiple-factor Design) 332.2.3 D-optimal Design 382.3 Strategies for Compound Selection 402.4 High Throughput Experiments 512.5 Summary 53References 543 Data Pre-treatment and Variable Selection 573.1 Introduction 573.2 Data Distribution 583.3 Scaling 603.4 Correlations 623.5 Data Reduction 633.6 Variable Selection 673.7 Summary 72References 734 Data Display 754.1 Introduction 754.2 Linear Methods 774.3 Nonlinear Methods 944.3.1 Nonlinear Mapping 944.3.2 Self-organizing Map 1054.4 Faces, Flowerplots and Friends 1104.5 Summary 113References 1165 Unsupervised Learning 1195.1 Introduction 1195.2 Nearest-neighbour Methods 1205.3 Factor Analysis 1255.4 Cluster Analysis 1355.5 Cluster Significance Analysis 1405.6 Summary 143References 1446 Regression Analysis 1456.1 Introduction 1456.2 Simple Linear Regression 1466.3 Multiple Linear Regression 1546.3.1 Creating Multiple Regression Models 1596.3.1.1 Forward Inclusion 1596.3.1.2 Backward Elimination 1616.3.1.3 Stepwise Regression 1636.3.1.4 All Subsets 1646.3.1.5 Model Selection by Genetic Algorithm 1656.3.2 Nonlinear Regression Models 1676.3.3 Regression with Indicator Variables 1696.4 Multiple Regression: Robustness, Chance Effects, the Comparison of Models and Selection Bias 1746.4.1 Robustness (Cross-validation) 1746.4.2 Chance Effects 1776.4.3 Comparison of Regression Models 1786.4.4 Selection Bias 1806.5 Summary 183References 1847 Supervised Learning 1877.1 Introduction 1877.2 Discriminant Techniques 1887.2.1 Discriminant Analysis 1887.2.2 SIMCA 1957.2.3 Confusion Matrices 1987.2.4 Conditions and Cautions for Discriminant Analysis 2017.3 Regression on Principal Components and PLS 2027.3.1 Regression on Principal Components 2037.3.2 Partial Least Squares 2067.3.3 Continuum Regression 2117.4 Feature Selection 2147.5 Summary 216References 2178 Multivariate Dependent Data 2198.1 Introduction 2198.2 Principal Components and Factor Analysis 2218.3 Cluster Analysis 2308.4 Spectral Map Analysis 2338.5 Models with Multivariate Dependent and Independent Data 2388.6 Summary 246References 2479 Artificial Intelligence and Friends 2499.1 Introduction 2509.2 Expert Systems 2519.2.1 LogP Prediction 2529.2.2 Toxicity Prediction 2619.2.3 Reaction and Structure Prediction 2689.3 Neural Networks 2739.3.1 Data Display Using ANN 2779.3.2 Data Analysis Using ANN 2809.3.3 Building ANN Models 2879.3.4 Interrogating ANN Models 2929.4 Miscellaneous AI Techniques 2959.5 Genetic Methods 3019.6 Consensus Models 3039.7 Summary 304References 30510 Molecular Design 30910.1 The Need for Molecular Design 30910.2 What is QSAR/QSPR? 31010.3 Why Look for Quantitative Relationships? 32110.4 Modelling Chemistry 32310.5 Molecular Fields and Surfaces 32510.6 Mixtures 32710.7 Summary 329References 330Index 333