ñòð. 122 |

aggression, behavior-based

147â€“148

variables, 18

ad hoc questions

AI (artificial intelligence), 15

behavior-based variables, 585

algorithms, recursive, 173

business opportunities,

alphas, decision trees, 188

identifying, 27

American Express

hypothesis testing, 50â€“51

as information broker, 16

additive facts, OLAP, 501

orders, market based analysis, 292

addresses, geographical resources,

555â€“556

615

616 Index

analysis sensitivity, 247â€“248

differential response, 107â€“108 sequential, 318â€“319

link analysis statistical

acyclic graphs, 331

acuity of testing, marketing

authorities, 333â€“334

campaign approaches, 147â€“148

candidates, 333

business data versus scientific

case study, 343â€“346

data, 159

classification, 9

censored data, 161

communities of interest, graphs, 346

Central Limit Theorem, 129â€“130

cyclic graphs, 330â€“331

chi-square tests, 149â€“153

data, as graphs, 340

confidence intervals, marketing

directed graphs, 330

campaign approaches, 146

discussed, 321

continuous variables, 137â€“138

edges, graphs, 322

correlation ranges, 139

fax machines, 337â€“341

cross-tabulations, 136

graph-coloring algorithm, 340â€“341

density function, 133

Hamiltonian path, graphs, 328

as disciplinary technique, 123

hubs, 332â€“334

discrete values, 127â€“131

Kleinberg algorithm, 332â€“333

experimentation, 160â€“161

nodes, graphs, 322

field values, 128

planar graphs, 323

histograms and, 127

root sets, 333

mean values, 137

search programs, 331

median values, 137

stemming, 333

mode values, 137

traveling salesman problem,

multiple comparisons, 148â€“149

graphs, 327â€“329

normal distribution, 130â€“132

vertices, graphs, 322

null hypothesis and, 125â€“126

weighted graphs, 322, 324

probabilities, 133â€“135

market based

proportion, standard error of,

differentiation, 289

marketing campaign

discussed, 287

approaches, 139â€“141

geographic attributes, 293

p-values, 126

item popularity, 293

q-values, 126

item sets, 289

range values, 137

market basket data, 51, 289â€“291

regression ranges, 139

marketing interventions, tracking,

sample sizes, marketing campaign

293â€“294

approaches, 145

order characteristics, 292

sample variation, 129

products, clustering by usage,

standard deviation, 132, 138

294â€“295

standardized values, 129â€“133

purchases, 289

sum of values, 137â€“138

support, 301

time series analysis, 128â€“129

telecommunications customers, 288

truncated data, 162

time attributes, 293

Index 617

sequential analysis, 318â€“319

variance, 138

for store comparisons, 315â€“316

z-values, 131, 138

trivial rules, 297

survival

virtual items, 307

attrition, handling different types

assumptions, validation, 67

of, 412â€“413

attrition

customer relationships, 413â€“415

discussed, 17

estimation tasks, 10

forced, 118

forecasting, 415â€“416

future, 49

time series

proof-of-concept projects, 599

neural networks, 244â€“247

survival analysis, 412â€“413

non-time series data, 246

audio, binary data, 557

SQL data, 572â€“573

authorities, link analysis, 333â€“334

statistics, 128â€“129

automated systems

of variance, 124

neural networks, 213

analysts, responsibilities of, 492â€“493

transaction processing systems, 3â€“4

analytic efforts, wasted time, 27

automatic cluster detection

AND value, neural networks, 222

agglomerative clustering, 368â€“370

angles, between vectors, 361â€“362

case study, 374â€“378

anonymous versus identified

categorical variables, 359

transactions, association rules, 308

centroid distance, 369

application programming interface

complete linkage, 369

(API), 535

data preparation, 363â€“365

architecture, data mining, 528â€“532

dimension, 352

artificial intelligence (AI), 15

directed clustering, 372

assessing models

discussed, 12, 91, 351

classifiers and predictors, 79

distance and similarity, 359â€“363

ñòð. 122 |