Data Mining. Concepts and Techniques, 3rd Edition
HAN 22-ind-673-708-9780123814791
Yüklə 7,95 Mb. Pdf görüntüsü
|
HAN 22-ind-673-708-9780123814791 2011/6/1 3:27 Page 700 #28 700 Index similarity (Continued) measuring, 65–78, 79 nominal attributes, 70 similarity measures, 447–448, 525–528 constraints on, 533 geodesic distance, 525–526 SimRank, 526–528 similarity searches, 587 in information networks, 594 in multimedia data mining, 596 simple random sample with replacement (SRSWR), 108 simple random sample without replacement (SRSWOR), 108 SimRank, 526–528, 539 computation, 527–528 random walk, 526–528 structural context, 528 simultaneous aggregation, 195 single-dimensional association rules, 17, 287 single-linkage algorithm, 460, 461 singular value decomposition (SVD), 587 skewed data balanced, 271 negatively, 47 positively, 47 wavelet transforms on, 102 slice operation, 148 small-world phenomenon, 592 smoothing, 112 by bin boundaries, 89 by bin means, 89 by bin medians, 89 for data discretization, 90 snowflake schema, 140 example, 141 illustrated, 141 star schema versus, 140 social networks, 524–525, 526–528 densification power law, 592 evolution of, 594 mining, 623 small-world phenomenon, 592 See also networks social science/social studies data mining, 613 soft clustering, 501 soft constraints, 534, 539 example, 534 handling, 536–537 space-filling curve, 58 sparse data, 102 sparse data cubes, 190 sparsest cuts, 539 sparsity coefficient, 579 spatial data, 14 spatial data mining, 595 spatiotemporal data analysis, 319 spatiotemporal data mining, 595, 623–624 specialized SQL servers, 165 specificity measure, 367 spectral clustering, 520–522, 539 effectiveness, 522 framework, 521 steps, 520–522 speech recognition, 430 speed, classification, 369 spiral method, 152 split-point, 333, 340, 342 splitting attributes, 333 splitting criterion, 333, 342 splitting rules. See attribute selection measures splitting subset, 333 SQL, as relational query language, 10 square-error function, 454 squashing function, 403 standard deviation, 51 example, 51 function of, 50 star schema, 139 example, 139–140 illustrated, 140 snowflake schema versus, 140 Star-Cubing, 204–210, 235 algorithm illustration, 209 bottom-up computation, 205 example, 207 for full cube computation, 210 ordering of dimensions and, 210 performance, 210 shared dimensions, 204–205 starnet query model, 149 example, 149–150 star-nodes, 205 star-trees, 205 compressed base table, 207 construction, 205 statistical data mining, 598–600 analysis of variance, 600 discriminant analysis, 600 factor analysis, 600 generalized linear models, 599–600 mixed-effect models, 600 quality control, 600
HAN 22-ind-673-708-9780123814791 2011/6/1 3:27 Page 701 #29 Index 701 regression, 599 survival analysis, 600 statistical databases (SDBs), 148 OLAP systems versus, 148–149 statistical descriptions, 24, 79 graphic displays, 44–45, 51–56 measuring the dispersion, 48–51 statistical hypothesis test, 24 statistical models, 23–24 of networks, 592–594 statistical outlier detection methods, 552, 553–560, 581 computational cost of, 560 for data analysis, 625 effectiveness, 552 example, 552 nonparametric, 553, 558–560 parametric, 553–558
statistical theory, in exceptional behavior disclosure, 291 statistics, 23 inferential, 24 predictive, 24 StatSoft, 602, 603 stepwise backward elimination, 105 stepwise forward selection, 105 stick figure visualization, 61–63 STING, 479–481 advantages, 480–481 as density-based clustering method, 480 hierarchical structure, 479, 480 multiresolution approach, 481
stratified cross-validation, 371 stratified samples, 109–110 stream data, 598, 624 strong association rules, 272 interestingness and, 264–265 misleading, 265 Structural Clustering Algorithm for Networks (SCAN), 531–532 structural context-based similarity, 526 structural data analysis, 319 structural patterns, 282 structure similarity search, 592 structures as contexts, 575 discovery of, 318 indexing, 319 substructures, 243 Student’s t-test, 372 subcube queries, 216, 217–218 sub-itemset pruning, 263 subjective interestingness measures, 22 subject-oriented data warehouses, 126 subsequence, 589 matching, 587 subset checking, 263–264 subset testing, 250 subspace clustering, 448 frequent patterns for, 318–319 subspace clustering methods, 509, 510–511, 538 biclustering, 511 correlation-based, 511 examples, 538 subspace search methods, 510–511 subspaces bottom-up search, 510–511 cube space, 228–229 outliers in, 578–579 top-down search, 511 substitution matrices, 590 substructures, 243 sum of the squared error (SSE), 501 summary fact tables, 165 superset checking, 263 supervised learning, 24, 330 supervised outlier detection, 549–550 challenges, 550 support, 21 association rule, 21 group-based, 286 reduced, 285, 286 uniform, 285–286 support, rule, 245, 246 support vector machines (SVMs), 393, 408–415, 437
interest in, 408 maximum marginal hyperplane, 409, 412 nonlinear, 413–415 for numeric prediction, 408 with sigmoid kernel, 415 support vectors, 411 for test tuples, 412–413 training/testing speed improvement, 415 support vectors, 411, 437 illustrated, 411 SVM finding, 412 supremum distance, 73–74 surface web, 597 survival analysis, 600 SVMs. See support vector machines
HAN 22-ind-673-708-9780123814791 2011/6/1 3:27 Page 702 #30 702 Index symbolic sequences, 586, 588 applications, 589 sequential pattern mining in, 588–589 symmetric binary dissimilarity, 70 synchronous generalization, 175 T tables, 9 attributes, 9 contingency, 95 dimension, 136 fact, 165 tuples, 9 tag clouds, 64, 66 Tanimoto coefficient, 78 target classes, 15, 180 initial working relations, 177 prime relation, 175, 177 targeted marketing, 609 taxonomy formation, 20 technologies, 23–27, 33, 34 telecommunications industry, 611 temporal data, 14 term-frequency vectors, 77 cosine similarity between, 78 sparse, 77 table, 77 terminating conditions, 404 test sets, 330 test tuples, 330 text data, 14 text mining, 596–597, 624 theoretical foundations, 600–601, 625 three-layer neural networks, 399 threshold-moving approach, 385 tilted time windows, 598 timeliness, data, 85 time-series data, 586, 587 cyclic movements, 588 discretization and, 590 illustrated, 588 random movements, 588 regression analysis, 587–588 seasonal variations, 588 shapelets method, 590 subsequence matching, 587 transformation into aggregate approximations, 587
trend analysis, 588 trend or long-term movements, 588 time-series data analysis, 319 time-series forecasting, 588 time-variant data warehouses, 127 top-down design approach, 133, 151 top-down subspace search, 511 top-down view, 151 topic model, 26–27 top-k patterns/rules, 281 top-k queries, 225 example, 225–226 ranking cubes to answer, 226–227 results, 225 user-specified preference components, 225
top-k strategies comparison illustration, 311 summarized pattern, 311 traditional, 311 TrAdaBoost, 436 training
Bayesian belief networks, 396–397 data, 18
sets, 328 tuples, 332–333 transaction reduction, 255 transactional databases, 13 example, 13–14 transactions, components of, 13 transfer learning, 430, 435, 434–436, 438 applications, 435 approaches to, 436 heterogeneous, 436 negative transfer and, 436 target task, 435 traditional learning versus, 435 treemaps, 63, 65 trend analysis spatial, 595 in time-series data, 588 for time-series forecasting, 588 trends, data mining, 622–625, 626 triangle inequality, 73 trimmed mean, 46 trimodal, 47 true negatives, 365 true positives, 365 t-test, 372 tuples, 9 duplication, 98–99 negative, 364 partitioning, 334, 337 positive, 364 training, 332–333 two sample t-test, 373 HAN 22-ind-673-708-9780123814791 2011/6/1 3:27 Page 703 #31 Index 703 two-layer neural networks, 399 two-level hash index structure, 264
ubiquitous data mining, 618–620, 625 uncertainty sampling, 433 undersampling, 384, 386 example, 384–385 uniform support, 285–286 unimodal, 47 unique rules, 92 univariate distribution, 40 univariate Gaussian mixture model, 504 univariate outlier detection, 554–555 unordered attributes, 103 unordered rules, 358 unsupervised learning, 25, 330, 445, 490 clustering as, 25, 445, 490 example, 25 supervised learning versus, 330 unsupervised outlier detection, 550 assumption, 550 clustering methods acting as, 551 upper approximation, 427 user interaction, 30–31 V values
exception, 234 expected, 97, 234 missing, 88–89 residual, 234 in rules or patterns, 281 variables grouping, 231 predicate, 295 predictor, 105 response, 105 variance, 51, 98 example, 51 function of, 50 variant graph patterns, 591 version space, 433 vertical data format, 260 example, 260–262 frequent itemset mining with, 259–262, 272 video data analysis, 319 virtual warehouses, 133 visibility graphs, 537 visible points, 537 visual data mining, 602–604, 625 data mining process visualization, 603 data mining result visualization, 603 data visualization, 602–603 as discipline integration, 602 illustrations, 604–607 interactive, 604, 607 as mining trend, 624 Viterbi algorithm, 591 W warehouse database servers, 131 warehouse refresh software, 151 waterfall method, 152 wavelet coefficients, 100 wavelet transforms, 99, 100–102 discrete (DWT), 100–102 for multidimensional data, 102 on sparse and skewed data, 102 web directories, 28 web mining, 597, 624 content, 597 as mining trend, 624 structure, 597–598 usage, 598 web search engines, 28, 523–524 web-document classification, 435 weight arithmetic mean, 46 weighted Euclidean distance, 74 Wikipedia, 597 WordNet, 597 working relations, 172 initial, 168, 169 World Wide Web (WWW), 1–2, 4, 14 Worlds-with-Worlds, 63, 64 wrappers, 127 Z z-score normalization, 114–115 Document Outline
Yüklə 7,95 Mb. Dostları ilə paylaş:
|