testing (continued) KS (Kolmogorov-Smirnov) tests, 101 preclassified tests, 79 test groups, marketing

campaigns, 106 test sets out of time tests, 72 uses for, 52 time

attributes, market based

analysis, 293 and dates, interval variables, 551 dependency, prospecting and, 160 frames, customer signatures, 562 series analysis neural networks, 244-247 non-time series data, 246 SQL data, 572-573 statistics, 128-129 training sets coverage of values, 232 MBR (memory-based reasoning),

263-264 model sets, partitioning, 71 optimization as, 230 uses for, 52 transaction data, OLAP, 476-477 transaction processing systems, customer relationship management, 3-4 transactional records, 574 transactors, behavior-based

variables, 580 transfer function, neural

networks, 223 TRANS MASTER file, customer

signatures, 559 traveling salesman problem, graph

theory, 327-329 trends, capturing, 75 triangle inequality, distance

function, 272 trivial rules, association rules, 297 truncated data, statistics, 162

truncated mean lifetime value,

retention, 389 truthful learning sources, 48-50 two-tailed distribution, 134

undirected data mining

affinity grouping, 57

clustering, 57

discussed, 7 uniform distribution, statistics, 132 uniform product code (UPC), 555 UNIT MASTER file, customer

signatures, 559 unordered lists, 239 unsupervised learning, 57 untruthful learning sources, 44-48 UPC (uniform product code), 555 UPS, transaction processing

systems, 3-4 up-selling

customer relationships, 467

marketing campaigns, 111, 115-116 U.S. Census Bureau Web site, 94 usage stimulation marketing

campaigns, 111 user roles, data transformation, 58-60

validation assumptions, 67 neural networks, 218 validation sets model sets, partitioning, 71 test sets, partitioning, 71 uses for, 52 value added-services, predication

tasks, 10 valued outcomes, estimation, 9 values

comparing with descriptions, 65 with meaning, data correction, 74 missing, 590-591

Team-Fly®

variables

data selection, 63-64

variable selection problems, neural networks, 233 variance

analysis of, 124

defined, 81

neural networks, 199

reduction in, splits, decision trees, 183

standard deviation and, 138

statistics, 138 variations, percent, 105 vectors, angles between, 361-362 vendor credibility, 537 virtual items, association rules, 307 virtuous cycle

action tasks, 30

business opportunities, identifying, 27-28

data transformation, 28-30

discussed, 28

results, measuring, 30-32

stages of, 26 visualization tools, data exploration, 65 voice node, fax machines, 341 voice recognition, free text

resources, 556 voluntary churn, 118-119, 521 voting

positive ratings, 284

weighted, 281-282

warehouses, searching data in, 61-62 warranty claims data, useful data

sources, 60 web crawlers, spiders, 331 Web pages

classification, 9

useful data sources, 60

Web servers cookies, 109

transaction processing systems, 3 Web sites customer response to marketing

campaigns, tracking, 109 RuleQuest, 190 U.S. Census Bureau, 94 weight columns, 548 weighted graphs, graph theory,

322, 324 weighted voting, 281-282 weighting, automatic cluster

detection, 363-365 welcome periods, loyalty

programs, 518 well-defined distance, distance

function, 271 winback approach, customer

relationships, 470 wireless communications industries business opportunities, identifying, 34-35

MOU (minutes of use), 38

rate plans, finding appropriate, 7

women, differential response analysis and, 107

word-of-mouth advertising, 283

zip codes

as categorical value, 239

distance function, 276-277 zone boundaries, adjusting, using automatic cluster detection, 380 z-scores, 551

z-values, statistics, 131, 138