A good E-book: <<Cluster_and_Classification_Techniques_for_the_Biosciences>>, Cambridge University Press,by Alan H. Fielding, 2007 pp260

1 Introduction 1
1.1 Background 1
1.2 Book structure 2
1.3 Classification 2
1.4 Clustering 3
1.5 Structures in data 3
1.6 Glossary 5
1.7 Recommended reading and other resources 10
2 Exploratory data analysis 12
2.1 Background 12
2.2 Dimensionality 13
2.3 Goodness of fit testing 14
2.4 Graphical methods 15
2.5 Variance-based data projections 16
2.6 Distance-based data projections 29
2.7 Other projection methods 32
2.8 Other methods 36
2.9 Data dredging 38
2.10 Example EDA analysis 38
3 Cluster analysis 46
3.1 Background 46
3.2 Distance and similarity measures 48
3.3 Partitioning methods 55
3.4 Agglomerative hierarchical methods 58
3.5 How many groups are there? 62
3.6 Divisive hierarchical methods 65

3.7 Two-way clustering and gene shaving 66
3.8 Recommended reading 67
3.9 Example analyses 68
4 Introduction to classification 78
4.1 Background 78
4.2 Black-box classifiers 81
4.3 Nature of a classifier 82
4.4 No-free-lunch 85
4.5 Bias and variance 86
4.6 Variable (feature) selection 87
4.7 Multiple classifiers 92
4.8 Why do classifiers fail? 94
4.9 Generalisation 95
4.10 Types of classifier 96
5 Classification algorithms 1 97
5.1 Background 97
5.2 Naı¨ve Bayes 99
5.3 Discriminant analysis 100
5.4 Logistic regression 117
5.5 Discriminant analysis or logistic regression? 128
5.6 Generalised additive models 130
5.7 Summary 136
6 Other classification methods 137
6.1 Background 137
6.2 Decision trees 137
6.3 Support vector machines 154
6.4 Artificial neural networks 156
6.5 Genetic algorithms 170
6.6 Others 175
6.7 Where next? 177
7 Classification accuracy 179
7.1 Background 179
7.2 Appropriate metrics 180
7.3 Binary accuracy measures 180
7.4 Appropriate testing data 183
7.5 Decision thresholds 186
7.6 Example 187
7.7 ROC plots 190
7.8 Incorporating costs 194
7.9 Comparing classifiers 196
7.10 Recommended reading 199