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 684 #12 684 Index data transformation (Continued) attribute construction, 112 in back-end tools/utilities, 134 concept hierarchy generation, 112, 120 discretization, 111, 112, 120 normalization, 112, 113–115, 120 smoothing, 112 strategies, 112–113
data types complex, 166 complex, mining, 585–598 for data mining, 8 data validation, 592–593 data visualization, 56–65, 79, 602–603 complex data and relations, 64–65 geometric projection techniques, 58–60 hierarchical techniques, 63–64 icon-based techniques, 60–63 mining process, 603 mining result, 603, 605 pixel-oriented techniques, 57–58 in science applications, 613 summary, 65 tag clouds, 64, 66 techniques, 39–40 data warehouses, 10–13, 26, 33, 125–185 analytical processing, 153 back-end tools/utilities, 134, 178 basic concepts, 125–135 bottom-up design approach, 133, 151–152 business analysis framework for, 150 business query view, 151 combined design approach, 152 data mart, 132, 142 data mining, 154 data source view, 151 design process, 151 development approach, 133 development tools, 153 dimensions, 10 enterprise, 132 extractors, 151 fact constellation, 141–142 for financial data, 608 framework illustration, 11 front-end client layer, 132 gateways, 131 geographic, 595 implementation, 156–165 information processing, 153 integrated, 126 metadata, 134–135 modeling, 10, 135–150 models, 132–134 multitier, 134 multitiered architecture, 130–132 nonvolatile, 127 OLAP server, 132 operational database systems versus, 128–129 planning and analysis tools, 153 retail industry, 609–610 in science applications, 612 snowflake schema, 140–141 star schema, 139–140 subject-oriented, 126 three-tier architecture, 131, 178 time-variant, 127 tools, 11 top-down design approach, 133, 151 top-down view, 151 update-driven approach, 128 usage for information processing, 153 view, 151 virtual, 133 warehouse database server, 131 database management systems (DBMSs), 9 database queries. See queries databases, 9 inductive, 601 relational. See relational databases research, 26 statistical, 148–149 technology evolution, 3 transactional, 13–15 types of, 32 web-based, 4 data/pattern analysis. See data mining DBSCAN, 471–473 algorithm illustration, 474 core objects, 472 density estimation, 477 density-based cluster, 472 density-connected, 472, 473 density-reachable, 472, 473 directly density-reachable, 472 neighborhood density, 471
DDPMine, 422 decimal scaling, normalization by, 115 decision tree analysis, discretization by, 116 decision tree induction, 330–350, 385 algorithm differences, 336 algorithm illustration, 333
HAN 22-ind-673-708-9780123814791 2011/6/1 3:27 Page 685 #13 Index 685 attribute selection measures, 336–344 attribute subset selection, 105 C4.5, 332 CART, 332 CHAID, 343 gain ratio, 340–341 Gini index, 332, 341–343 ID3, 332 incremental versions, 336 information gain, 336–340 multivariate splits, 344 parameters, 332 scalability and, 347–348 splitting criterion, 333 from training tuples, 332–333 tree pruning, 344–347, 385 visual mining for, 348–350 decision trees, 18, 330 branches, 330 illustrated, 331 internal nodes, 330 leaf nodes, 330 pruning, 331, 344–347 root node, 330 rule extraction from, 357–359 deep web, 597 default rules, 357 DENCLUE, 476–479 advantages, 479 clusters, 478 density attractor, 478 density estimation, 476 kernel density estimation, 477–478 kernels, 478
dendrograms, 460 densification power law, 592 density estimation, 476 DENCLUE, 477–478 kernel function, 477–478 density-based methods, 449, 471–479, 491 DBSCAN, 471–473 DENCLUE, 476–479 object division, 449 OPTICS, 473–476 STING as, 480 See also cluster analysis density-based outlier detection, 564–567 local outlier factor, 566–567 local proximity, 564 local reachability density, 566 relative density, 565 descendant cells, 189 descriptive mining tasks, 15 DIANA (Divisive Analysis), 459, 460 dice operation, 148 differential privacy, 622 dimension tables, 136 dimensional cells, 189 dimensionality reduction, 86, 99–100, 120 dimensionality reduction methods, 510, 519–522, 538 list of, 587 spectral clustering, 520–522 dimension/level application of, 297 constraints, 294 dimensions, 10, 136 association rule, 281 cardinality of, 159 concept hierarchies and, 142–144 in multidimensional view, 33 ordering of, 210 pattern, 281 ranking, 225 relevance analysis, 175 selection, 225 shared, 204 See also data warehouses direct discriminative pattern mining, 422 directed acyclic graphs, 394–395 discernibility matrix, 427 discovery-driven exploration, 231–234, 235 discrepancy detection, 91–93 discrete attributes, 44 discrete Fourier transform (DFT), 101, 587 discrete wavelet transform (DWT), 100–102, 587
discretization, 112, 120 by binning, 115 by clustering, 116 by correlation analysis, 117 by decision tree analysis, 116 by histogram analysis, 115–116 techniques, 113 discriminant analysis, 600 discriminant rules, 16 discriminative frequent pattern-based classification, 416, 419–422, 437 basis for, 419 feature generation, 420 feature selection, 420–421 framework, 420–421 learning of classification model, 421 Yüklə 7,95 Mb. Dostları ilə paylaş:
|