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 686 #14 686 Index dispersion of data, 44, 48–51 dissimilarity asymmetric binary, 71 between attributes of mixed type, 76–77 between binary attributes, 71–72 measuring, 65–78, 79 between nominal attributes, 69 on numeric data, 72–74 between ordinal attributes, 75 symmetric binary, 70–71 dissimilarity matrix, 67, 68 data matrix versus, 67–68
as one-mode matrix, 68 distance measures, 461–462 Euclidean, 72–73 Manhattan, 72–73 Minkowski, 73 supremum, 73–74 types of, 72 distance-based cluster analysis, 445 distance-based outlier detection, 561–562 nested loop algorithm, 561, 562
distributed data mining, 615, 624 distributed privacy preservation, 622 distributions boxplots for visualizing, 49–50 five-number summary, 49 distributive measures, 145 Divisive Analysis (DIANA), 459, 460 divisive hierarchical method, 459 agglomerative hierarchical clustering versus, 459–460 DIANA, 459, 460 DNA chips, 512 document classification, 430 documents language model, 26 topic model, 26–27 drill-across operation, 148 drill-down operation, 11, 146–147 drill-through operation, 148 dynamic itemset counting, 256
eager learners, 423, 437 Eclat (Equivalence Class Transformation) algorithm, 260, 272
e-commerce, 609 editing method, 425 efficiency Apriori algorithm, 255–256 backpropagation, 404 data mining algorithms, 31 elbow method, 486 email spam filtering, 435 engineering applications, 613 ensemble methods, 378–379, 386 bagging, 379–380 boosting, 380–382 for class imbalance problem, 385 random forests, 382–383 types of, 378, 386 enterprise warehouses, 132 entity identification problem, 94 entity-relationship (ER) data model, 9, 139 epoch updating, 404 equal-frequency histograms, 107, 116 equal-width histograms, 107, 116 equivalence classes, 427 error rates, 367 error-correcting codes, 431–432 Euclidean distance, 72 mathematical properties, 72–73 weighted, 74
evaluation metrics, 364–370 evolution, of database system technology, 3–5 evolutionary searches, 579 exception-based, discovery-driven exploration, 231–234, 235 exceptions, 231 exhaustive rules, 358 expectation-maximization (EM) algorithm, 505–508, 538 expectation step (E-step), 505 fuzzy clustering with, 505–507 maximization step (M-step), 505 for mixture models, 507–508 for probabilistic model-based clustering, 507–508
steps, 505 See also probabilistic model-based clustering expected values, 97 cell, 234 exploratory data mining. See multidimensional data mining extraction data, 134 rule, from decision tree, 357–359 extraction/transformation/loading (ETL) tools, 93 extractors, 151 HAN 22-ind-673-708-9780123814791 2011/6/1 3:27 Page 687 #15 Index 687 F fact constellation, 141 example, 141–142 illustrated, 142 fact tables, 136 summary, 165 factor analysis, 600 facts, 136 false negatives, 365 false positives, 365 farthest-neighbor clustering algorithm, 462 field overloading, 92 financial data analysis, 607–609 credit policy analysis, 608–609 crimes detection, 609 data warehouses, 608 loan payment prediction, 608–609 targeted marketing, 609 FindCBLOF algorithm, 569–570 five-number summary, 49 fixed-width clustering, 570 FOIL, 359, 363, 418 Forest-RC, 383 forward algorithm, 591 FP-growth, 257–259, 272 algorithm illustration, 260 example, 257–258 performance, 259 FP-trees, 257 condition pattern base, 258 construction, 257–258 main memory-based, 259 mining, 258, 259 Frag-Shells, 212, 213 fraudulent analysis, 610–611 frequency patterns approximate, 281, 307–312 compressed, 281, 307–312 constraint-based, 281 near-match, 281 redundancy-aware top-k, 281 top-k, 281 frequent itemset mining, 18, 272, 282 Apriori algorithm, 248–253 closed patterns, 262–264 market basket analysis, 244–246 max patterns, 262–264 methods, 248–264 pattern-growth approach, 257–259 with vertical data format, 259–262, 272 frequent itemsets, 243, 246, 272 association rule generation from, 253, 254 closed, 247, 248, 262–264, 308 finding, 247 finding by confined candidate generation, 248–253
maximal, 247, 248, 262–264, 308 subsets, 309 frequent pattern mining, 279 advanced forms of patterns, 320 application domain-specific semantics, 282 applications, 317–319, 321 approximate patterns, 307–312 classification criteria, 280–283 colossal patterns, 301–307 compressed patterns, 307–312 constraint-based, 294–301, 320 data analysis usages, 282 for data cleaning, 318 direct discriminative, 422 high-dimensional data, 301–307 in high-dimensional space, 320 in image data analysis, 319 for indexing structures, 319 kinds of data and features, 282 multidimensional associations, 287–289 in multilevel, multidimensional space, 283–294 multilevel associations, 283–294 in multimedia data analysis, 319 negative patterns, 291–294 for noise filtering, 318 Pattern-Fusion, 302–307 quantitative association rules, 289–291 rare patterns, 291–294 in recommender systems, 319 road map, 279–283 scalable computation and, 319 scope of, 319–320 in sequence or structural data analysis, 319 in spatiotemporal data analysis, 319 for structure and cluster discovery, 318 for subspace clustering, 318–319 in time-series data analysis, 319 top-k, 310 in video data analysis, 319
frequent pattern-based classification, 415–422, 437 associative, 415, 416–419 discriminative, 416, 419–422 framework, 422 frequent patterns, 17, 243 abstraction levels, 281 association rule mapping, 280 basic, 280
Yüklə 7,95 Mb. Dostları ilə paylaş:
|