The present work is an introductory text in statistics, addressed to researchers and students in the field of material science. It aims to give readers a basic knowledge of how statistical reasoning is used in this field, by improving their knowledge of statistical tools and helping them to carry out statistical analyses and interpret the results. It also focuses on establishing a consistent multivariate workflow starting from a correct design of the experiment followed by a multivariate analysis process.
Cover 1
Title Page 2
Copyright Page 3
Preface 4
Acknowledgements 8
Contents 10
PART I STATISTICS BASICS 16
1. Statistics Basics 17
1.1 Introduction 17
1.1.1 Data-set PLA 17
1.1.2 Data-set FGO 17
1.2 Samples and Variables 18
1.3 Errors 21
1.4 Initial Data Analysis 22
1.4.1 Significant Digits 22
1.4.2 Stripcharts, Stem-and-leaf Displays, and Histograms 23
1.5 Mode, Median, Mean, Variance, and Standard Deviation 28
1.5.1 The Median 28
1.5.2 Mode 30
1.5.3 Mean 30
1.5.4 A Visual Comparison of Mean, Median, and Mode 31
1.5.5 The Range 32
1.5.6 Quartile and Interquartile Range 32
1.5.7 Variance 33
1.5.8 The Standard Deviation 34
1.5.9 Distributions 34
1.6 Z-score 36
1.6.1 Box and Whiskers Plot 36
1.7 Error Propagation and Uncertainty 38
1.8 Normality Tests 40
1.9 Significance Tests 41
1.9.1 Outliers 42
1.9.2 Q-test 42
1.9.3 Cochran Test 42
1.10 T-test 43
1.11 F-test 45
1.12 One-way Analysis of Variance ANOVA 46
1.13 Two-way Analysis of Variance ANOVA 47
1.13.1 Two way ANOVA with Interaction 47
1.14 Type I, II, and III Errors 50
1.15 Bootstrap 51
1.15.1 Two-sample Problems: Comparing Means or Median? 51
1.16 An Example of Non-normal Distribution 52
1.17 About Visual Representation of Data 54
1.18 FAQ 54
1.18.1 Additional Data-set and Exercises 55
1.18.2 Remarks 56
1.18.3 Suggested Essential Literature 56
Bibliography 58
PART II ESSENTIAL MULTIVARIATE STATISTICS 60
2. Design of Experiment 61
2.1 Introduction 61
2.2 Randomization 62
2.3 Data-set OPT Cables 64
2.4 One Variable at a Time Design 64
2.5 Factorial Design 65
2.6 Regression Model Representations 67
2.6.1 Factorial Model Including Three Replicates in the Center 68
2.6.2 Model with More than Two Levels for each Factor 72
2.7 Data-set EMAGMA, An Example of DoE with Three Factors 74
2.7.1 Workflow using OVAT 75
2.7.2 Factorial Design 23 76
2.7.3 Factorial Design 2K 81
2.7.4 Fractional Factorial Design 2K-1 81
2.7.5 On Graphical Representation of Factorials with Four Factors 82
2.8 Mixture Design 83
2.8.1 Data Set HIPS 84
2.9 Design of Experiments Matrix vs Real Experiments Performed 86
2.9.1 Mixture Design in Constrained Region 86
2.9.2 Data Set CPCB 86
2.10 Other Designs 91
2.11 FAQ 92
2.11.1 Exercises 93
2.11.2 Remarks 94
2.11.3 Suggested Essential Literature 95
Bibliography 96
3. Pattern Recognition 98
3.1 Introduction 98
3.2 Variable Correlation 99
3.2.1 Datasaurus 102
3.3 Principal Component Analysis 103
3.3.1 Centering and Scaling 105
3.3.2 Algorithms for PCA 106
3.3.3 Data-set ELE: Example of PCA Applied to a Data-set Obtained Via Electrophoresis Characterization 107
3.3.4 Data-set ASPHALT: An Application of PCA to ATR-FTIR Spectroscopy 112
3.3.5 Data-set PCAMIX: PCA Applied to Binary Chemical Mixtures at Trace Levels 119
3.3.6 Cluster Analysis 123
3.3.7 Dendrograms 127
3.3.8 K-means Method 128
3.3.9 Discriminant Analysis 131
3.3.10 Soft Independent Modelling of Class Analogy 132
3.3.11 Artificial Neural Networks 138
3.3.12 Other Methodologies 139
3.3.13 Q.A. 139
3.3.14 Exercises 141
3.3.15 Remarks 141
3.3.16 Suggested Essential Literature 141
Bibliography 142
4. Calibration 144
4.1 Introduction 144
4.2 Univariate Calibration 144
4.3 Univariate Calibration, Data-set Concrete 146
4.3.1 Bivariate Models 147
4.4 Multivariate Calibration 153
4.4.1 Principal Component Regression 153
4.4.2 An Example of Multivariate Regression using the Gasoline Data Set 154
4.4.3 Partial Least Squares 157
4.5 Other Regression Methodologies 161
4.5.1 NWAY Methodologies 161
4.5.2 A Short History of Partial Least Squares 164
4.5.3 Q.A. 164
4.5.4 Essential References 165
Bibliography 166
5. Case Studies 167
5.1 Fast Fabrication of ZnO Superhydrophobic Surfaces without Chemical Post-treatment: Investigation of Important Parameters using Taguchi Mixed Level Design L8 (41 23) 168
5.2 Introduction 168
5.3 Materials and Methods 169
5.3.1 Materials 169
5.3.2 Design of Experiments (DOE) 170
5.4 Sample Preparation 171
5.4.1 Characterization 171
5.5 Results and Discussion 171
5.5.1 DOE Analysis 171
5.5.2 XRD Results 174
5.5.3 SEM Results 175
5.5.4 ATR-FTIR Analysis 177
5.6 Summary 179
Bibliography 180
5.7 An Example of Evolutionary Design of Experiment: Prediction of the Aging of Polymers 184
5.8 Introduction 184
5.8.1 Evolutionary Design of Experiment for Accelerated Aging Tests 186
5.9 Prediction of Rubber Aging by Accelerated Aging Tests 189
5.9.1 Successive Bayesian Estimation 192
5.10 Results and Discussion 193
5.11 Conclusions 197
Bibliography 199
5.12 Principal Component Analysis Applied to the Study of the Behavior of Steel Corrosion Inhibitors 200
5.13 Introduction 200
5.14 Materials and Methods 201
5.14.1 Samples Preparation 201
5.14.2 Chemical Speciation Equilibrium of Inhibitors 201
5.15 Electrode Preparation 202
5.16 Electrochemical Techniques 203
5.16.1 Zero Current Potential and Potentiodynamic Polarisation Measurement 204
5.17 Cyclic Voltammetry 204
5.18 Data Management Multivariate Analysis 204
5.19 Results and Discussion 204
5.19.1 Open Circuit Potential (OCP) and Tafel Polarization Measurement 204
5.20 Multivariate Analysis 207
5.20.1 Principal Component Analysis 207
5.20.2 Calibration-validation Test 207
5.20.3 Cyclic Voltammetry Study 208
5.21 Conclusions 211
Bibliography 212
Appendices 214
A Software Workflow 214
A.1 Software 214
Bibliography 216
A.2 Chapter 1 217
A.3 Chapter 2 227
A.4 Chapter 3 239
A.5 Chapter 4 252
A.6 Appendix 266
A.7 Plackett-Burman 16 267
A.8 Statistical Tables 267
B A Short Refresher of Matrix Algebra 268
C Statistical Tables 274
D Design of Experiment Tables 281
D.1 Factorial Design 281
D.2 Placket Burman 286
Index 288
chemometrics;,univariate,statistics;,multivariate,statistics;,analytical,chemistry;,corrosion;,spectroscopy;,infrared
chemometrics,univariate statistics,multivariate statistics,analytical chemistry,corrosion,spectroscopy,infrared

lgli/Giorgio Luciano - Statistical and Multivariate Analysis in Material Science (2021, CRC Press).pdf

zlib/no-category/Giorgio Luciano/Statistical and Multivariate Analysis in Material Science_11560257.pdf

Ashgate Publishing Limited

Taylor & Francis Group

Taylor & Francis Ltd

Gower Publishing Ltd

United Kingdom and Ireland, United Kingdom

Boca Raton, 2018

2020

sources:
9781138196308

producers:
iTextSharpTM 5.5.13 ©2000-2018 iText Group NV (AGPL-version)

The present work is an introductory text in statistics, addressed to researchers and students in the field of material science. It aims to give the readers basic knowledge on how statistical reasoning is exploitable in this field, improving their knowledge of statistical tools and helping them to carry out statistical analyses and to interpret the results. It also focuses on establishing a consistent multivariate workflow starting from a correct design of experiment followed by a multivariate analysis process.

This book introduces univariate and multivariate statistics applied to material science using an easy-to-follow, non-mathematical approach. Case studies and tutorials are provided to assist readers with the application of the techniques described utilizing their own data. The book will be a valuable tool for scientists and advanced students specializing in material science, corrosion science, and chemometrics.

2021-02-10