Data Mining. Concepts and Techniques, 3rd Edition

HAN

22-ind-673-708-9780123814791

2011/6/1

3:27

Page 692

#20

692

Index

link mining, 594

link prediction, 594

load, in back-end tools/utilities, 134

loan payment prediction, 608–609

local outlier factor, 566–567

local proximity-based outliers, 564–565

logistic function, 402

log-linear models, 106

lossless compression, 100

lossy compression, 100

lower approximation, 427

M

machine learning, 24–26

active, 25

data mining similarities, 26

semi-supervised, 25

supervised, 24

unsupervised, 25

Mahalanobis distance, 556

majority voting, 335

Manhattan distance, 72–73

MaPle, 519

margin, 410

market basket analysis, 244–246, 271–272

example, 244

illustrated, 244

Markov chains, 591

materialization

full, 159, 179, 234

iceberg cubes, 319

no, 159

partial, 159–160, 192, 234

semi-offline, 226

max patterns, 280

max confidence

measure, 268, 272

maximal frequent itemsets, 247, 308

example, 248

mining, 262–264

shortcomings for compression, 308–309

maximum marginal hyperplane (MMH), 409

SVM finding, 412

maximum normed residual test, 555

mean, 39, 45

bin, smoothing by, 89

example, 45

for missing values, 88

trimmed, 46

weighted arithmetic, 45

measures, 145

accuracy-based, 369

algebraic, 145

all confidence

, 272

antimonotonic, 194

attribute selection, 331

categories of, 145

of central tendency, 39, 44, 45–47

correlation, 266

data cube, 145

dispersion, 48–51

distance, 72–74, 461–462

distributive, 145

holistic, 145

Kulczynski, 272

max confidence

, 272

of multidimensional databases, 146

null-invariant, 272

pattern evaluation, 267–271

precision, 368–369

proximity, 67, 68–72

recall, 368–369

sensitivity, 367

significance, 312

similarity/dissimilarity, 65–78

specificity, 367

median, 39, 46

bin, smoothing by, 89

example, 46

formula, 46–47

for missing values, 88

metadata, 92, 134, 178

business, 135

importance, 135

operational, 135

repositories, 134–135

metarule-guided mining

of association rules, 295–296

example, 295–296

metrics, 73

classification evaluation, 364–370

microeconomic view, 601

midrange, 47

MineSet, 603, 605

minimal interval size, 116

minimal spanning tree algorithm, 462

minimum confidence threshold, 18, 245

Minimum Description Length (MDL), 343–344

minimum support threshold, 18, 190

association rules, 245

count, 246

Minkowski distance, 73

min-max normalization, 114

missing values, 88–89

mixed-effect models, 600

---

HAN

22-ind-673-708-9780123814791

2011/6/1

3:27

Page 693

#21

Index

693

mixture models, 503, 538

EM algorithm for, 507–508

univariate Gaussian, 504

mode, 39, 47

example, 47

model selection, 364

with statistical tests of significance, 372–373

models, 18

modularity

of clustering, 530

use of, 539

MOLAP. See multidimensional OLAP

monotonic constraints, 298

motifs, 587

moving-object data mining, 595–596, 623–624

multiclass classification, 430–432, 437

all-versus-all (AVA), 430–431

error-correcting codes, 431–432

one-versus-all (OVA), 430

multidimensional association rules, 17, 283,

288, 320

hybrid-dimensional, 288

interdimensional, 288

mining, 287–289

mining with static discretization of quantitative

attributes, 288

with no repeated predicates, 288

See also association rules

multidimensional data analysis

in cube space, 227–234

in multimedia data mining, 596

spatial, 595

of top-k results, 226

multidimensional data mining, 11–13, 34 155–156,

179, 187, 227, 235

data cube promotion of, 26

dimensions, 33

example, 228–229

retail industry, 610

multidimensional data model, 135–146, 178

data cube as, 136–139

dimension table, 136

dimensions, 142–144

fact constellation, 141–142

fact table, 136

snowflake schema, 140–141

star schema, 139–140

multidimensional databases

measures of, 146

querying with starnet model, 149–150

multidimensional histograms, 108

multidimensional OLAP (MOLAP), 132, 164, 179

multifeature cubes, 227, 230, 235

complex query support, 231

examples, 230–231

multilayer feed-forward neural networks,

398–399

example, 405

illustrated, 399

layers, 399

units, 399

multilevel association rules, 281, 283, 284, 320

ancestors, 287

concept hierarchies, 285

dimensions, 281

group-based support, 286

mining, 283–287

reduced support, 285, 286

redundancy, checking, 287

uniform support, 285–286

multimedia data, 14

multimedia data analysis, 319

multimedia data mining, 596

multimodal, 47

multiple linear regression, 90, 106

multiple sequence alignment, 590

multiple-phase clustering, 458–459

multitier data warehouses, 134

multivariate outlier detection, 556

with Mahalanobis distance, 556

with multiple clusters, 557

with multiple parametric distributions, 557

with

χ

2

-static, 556

multiway array aggregation, 195, 235

for full cube computation, 195–199

minimum memory requirements, 198

must-link constraints, 533, 536

mutation operator, 426

mutual information, 315–316

mutually exclusive rules, 358

N

naive Bayesian classification, 385

class label prediction with, 353–355

functioning of, 351–352

nearest-neighbor clustering algorithm, 461

near-match patterns/rules, 281

negative correlation, 55, 56

negative patterns, 280, 283, 320

example, 291–292

mining, 291–294

negative transfer, 436

negative tuples, 364

negatively skewed data, 47

---