Applied data mining: statistical methods for business and industry

Applied Data Mining......Page 3
Contents......Page 7
Preface......Page 13
1.1 What is data mining?......Page 15
1.1.1 Data mining and computing......Page 17
1.1.2 Data mining and statistics......Page 19
1.2 The data mining process......Page 20
1.3 Software for data mining......Page 25
1.4 Organisation of the book......Page 26
1.4.2 Chapters 7 to 12: business cases......Page 27
1.5 Further reading......Page 28
Part I Methodology......Page 31
2 Organisation of the data......Page 33
2.1.1 The data warehouse......Page 34
2.1.2 The data webhouse......Page 35
2.2 Classification of the data......Page 36
2.3 The data matrix......Page 37
2.4 Frequency distributions......Page 39
2.4.1 Univariate distributions......Page 40
2.4.2 Multivariate distributions......Page 41
2.5 Transformation of the data......Page 43
2.6 Other data structures......Page 44
2.7 Further reading......Page 45
3 Exploratory data analysis......Page 47
3.1 Univariate exploratory analysis......Page 48
3.1.1 Measures of location......Page 49
3.1.3 Measures of heterogeneity......Page 51
3.1.4 Measures of concentration......Page 53
3.1.5 Measures of asymmetry......Page 55
3.1.6 Measures of kurtosis......Page 57
3.2 Bivariate exploratory analysis......Page 59
3.3 Multivariate exploratory analysis of quantitative data......Page 63
3.4 Multivariate exploratory analysis of qualitative data......Page 65
3.4.1 Independence and association......Page 67
3.4.2 Distance measures......Page 68
3.4.3 Dependency measures......Page 70
3.4.4 Model-based measures......Page 72
3.5 Reduction of dimensionality......Page 75
3.5.1 Interpretation of the principal components......Page 77
3.5.2 Application of the principal components......Page 79
3.6 Further reading......Page 80
4 Computational data mining......Page 83
4.1 Measures of distance......Page 84
4.1.1 Euclidean distance......Page 85
4.1.2 Similarity measures......Page 86
4.1.3 Multidimensional scaling......Page 88
4.2 Cluster analysis......Page 89
4.2.1 Hierarchical methods......Page 91
4.2.2 Evaluation of hierarchical methods......Page 95
4.2.3 Non-hierarchical methods......Page 97
4.3.1 Bivariate linear regression......Page 99
4.3.2 Properties of the residuals......Page 102
4.3.3 Goodness of fit......Page 104
4.3.4 Multiple linear regression......Page 105
4.4 Logistic regression......Page 110
4.4.1 Interpretation of logistic regression......Page 111
4.4.2 Discriminant analysis......Page 112
4.5 Tree models......Page 114
4.5.1 Division criteria......Page 117
4.5.2 Pruning......Page 119
4.6 Neural networks......Page 121
4.6.1 Architecture of a neural network......Page 123
4.6.2 The multilayer perceptron......Page 125
4.6.3 Kohonen networks......Page 131
4.7 Nearest-neighbour models......Page 133
4.8.1 Association rules......Page 135
4.8.2 Retrieval by content......Page 140
4.9 Further reading......Page 141
5.1 Uncertainty measures and inference......Page 143
5.1.1 Probability......Page 144
5.1.2 Statistical models......Page 146
5.1.3 Statistical inference......Page 151
5.2 Non-parametric modelling......Page 157
5.3 The normal linear model......Page 160
5.3.1 Main inferential results......Page 161
5.3.2 Application......Page 164
5.4 Generalised linear models......Page 168
5.4.1 The exponential family......Page 169
5.4.2 Definition of generalised linear models......Page 171
5.4.3 The logistic regression model......Page 177
5.4.4 Application......Page 178
5.5.1 Construction of a log-linear model......Page 181
5.5.2 Interpretation of a log-linear model......Page 183
5.5.3 Graphical log-linear models......Page 185
5.5.4 Log-linear model comparison......Page 188
5.5.5 Application......Page 189
5.6 Graphical models......Page 191
5.6.1 Symmetric graphical models......Page 192
5.6.2 Recursive graphical models......Page 196
5.6.3 Graphical models versus neural networks......Page 198
5.7 Further reading......Page 199
6 Evaluation of data mining methods......Page 201
6.1.1 Distance between statistical models......Page 202
6.1.2 Discrepancy of a statistical model......Page 204
6.1.3 The Kullback–Leibler discrepancy......Page 206
6.2 Criteria based on scoring functions......Page 207
6.3 Bayesian criteria......Page 209
6.4 Computational criteria......Page 211
6.5 Criteria based on loss functions......Page 214
6.6 Further reading......Page 218
Part II Business cases......Page 221
7.1 Objectives of the analysis......Page 223
7.2 Description of the data......Page 224
7.3 Exploratory data analysis......Page 226
7.4.1 Log-linear models......Page 229
7.4.2 Association rules......Page 232
7.5 Model comparison......Page 238
7.6 Summary report......Page 240
8.2 Description of the data......Page 243
8.3 Exploratory data analysis......Page 246
8.4.1 Sequence rules......Page 252
8.4.2 Link analysis......Page 256
8.4.3 Probabilistic expert systems......Page 258
8.4.4 Markov chains......Page 259
8.5 Model comparison......Page 264
8.6 Summary report......Page 266
9.2 Description of the data......Page 269
9.4.1 Cluster analysis......Page 272
9.4.2 Kohonen maps......Page 276
9.5 Model comparison......Page 278
9.6 Summary report......Page 285
10.2 Description of the data......Page 287
10.3 Exploratory data analysis......Page 289
10.4.1 Logistic regression models......Page 292
10.4.2 Radial basis function networks......Page 294
10.4.3 Classification tree models......Page 295
10.4.4 Nearest-neighbour models......Page 299
10.5 Model comparison......Page 300
10.6 Summary report......Page 304
11.1 Objectives of the analysis......Page 307
11.2 Description of the data......Page 308
11.3 Exploratory data analysis......Page 310
11.4.1 Logistic regression models......Page 313
11.4.2 Classification tree models......Page 317
11.5 Model comparison......Page 328
11.6 Summary report......Page 333
12.1 Objectives of the analysis......Page 337
12.2 Description of the data......Page 338
12.3 Exploratory data analysis......Page 341
12.4 Model building......Page 351
12.5 Model comparison......Page 361
12.6 Summary report......Page 364
Bibliography......Page 367
Index......Page 371