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 676 #4 676 Index biclusters, 511 with coherent values, 516 with coherent values on rows, 516 with constant values, 515 with constant values on columns, 515 with constant values on rows, 515 as submatrix, 515 types of, 515–516 bimodal, 47 bin boundaries, 89 binary attributes, 41, 79 asymmetric, 42, 70 as Boolean, 41 contingency table for, 70 dissimilarity between, 71–72 example, 41–42 proximity measures, 70–72 symmetric, 42, 70–71
binning
discretization by, 115 equal-frequency, 89 smoothing by bin boundaries, 89 smoothing by bin means, 89 smoothing by bin medians, 89 biological sequences, 586, 624 alignment of, 590–591 analysis, 590 BLAST, 590 hidden Markov model, 591 as mining trend, 624 multiple sequence alignment, 590 pairwise alignment, 590 phylogenetic tree, 590 substitution matrices, 590 bipartite graphs, 523 BIRCH, 458, 462–466 CF-trees, 462–463, 464, 465–466 clustering feature, 462, 463, 464 effectiveness, 465 multiphase clustering technique, 464–465
bitmap indexing, 160–161, 179 bitmapped join indexing, 163, 179 bivariate distribution, 40 BLAST. See Basic Local Alignment Search Tool BOAT. See Bootstrapped Optimistic Algorithm for Tree construction Boolean association rules, 281 Boolean attributes, 41 boosting, 380 accuracy, 382 AdaBoost, 380–382 bagging versus, 381–382 weight assignment, 381 bootstrap method, 371, 386 bottom-up design approach, 133, 151–152 bottom-up subspace search, 510–511 boxplots, 49 computation, 50 example, 50 five-number summary, 49 illustrated, 50 in outlier visualization, 555 BUC, 200–204, 235 for 3-D data cube computation, 200 algorithm, 202 Apriori property, 201 bottom-up construction, 201 iceberg cube construction, 201 partitioning snapshot, 203 performance, 204 top-down processing order, 200, 201 business intelligence (BI), 27 business metadata, 135 business query view, 151
C4.5, 332, 385 class-based ordering, 358 gain ratio use, 340 greedy approach, 332 pessimistic pruning, 345 rule extraction, 358
cannot-link constraints, 533 CART, 332, 385 cost complexity pruning algorithm, 345 Gini index use, 341 greedy approach, 332 See also decision tree induction case updating, 404 case-based reasoning (CBR), 425–426 challenges, 426 categorical attributes, 41 CBA. See Classification Based on Associations CBLOF. See cluster-based local outlier factor CELL method, 562, 563 cells, 10–11 aggregate, 189 ancestor, 189 base, 189 descendant, 189
HAN 22-ind-673-708-9780123814791 2011/6/1 3:27 Page 677 #5 Index 677 dimensional, 189 exceptions, 231 residual value, 234 central tendency measures, 39, 44, 45–47 mean, 45–46 median, 46–47 midrange, 47 for missing values, 88 models, 47 centroid distance, 108 CF-trees, 462–463, 464 nodes, 465 parameters, 464 structure illustration, 464 CHAID, 343 Chameleon, 459, 466–467 clustering illustration, 466 relative closeness, 467 relative interconnectivity, 466–467 See also hierarchical methods Chernoff faces, 60 asymmetrical, 61 illustrated, 62 ChiMerge, 117 chi-square test, 95 chunking, 195 chunks, 195 2-D, 197 3-D, 197
computation of, 198 scanning order, 197 CLARA. See Clustering Large Applications CLARANS. See Clustering Large Applications based upon Randomized Search class comparisons, 166, 175, 180 attribute-oriented induction for, 175–178
mining, 176 presentation of, 175–176 procedure, 175–176 class conditional independence, 350 class imbalance problem, 384–385, 386 ensemble methods for, 385 on multiclass tasks, 385 oversampling, 384–385, 386 threshold-moving approach, 385 undersampling, 384–385, 386 class label attributes, 328 class-based ordering, 357 class/concept descriptions, 15 classes, 15, 166 contrasting, 15 equivalence, 427 target, 15 classification, 18, 327–328, 385 accuracy, 330 accuracy improvement techniques, 377–385 active learning, 433–434 advanced methods, 393–442 applications, 327 associative, 415, 416–419, 437 automatic, 445 backpropagation, 393, 398–408, 437 bagging, 379–380 basic concepts, 327–330 Bayes methods, 350–355 Bayesian belief networks, 393–397, 436 boosting, 380–382 case-based reasoning, 425–426 of class-imbalanced data, 383–385 confusion matrix, 365–366, 386 costs and benefits, 373–374 decision tree induction, 330–350 discriminative frequent pattern-based, 437 document, 430 ensemble methods, 378–379 evaluation metrics, 364–370 example, 19 frequent pattern-based, 393, 415–422, 437 fuzzy set approaches, 428–429, 437 general approach to, 328 genetic algorithms, 426–427, 437 heterogeneous networks, 593 homogeneous networks, 593 IF-THEN rules for, 355–357 interpretability, 369
lazy learners, 393, 422–426 learning step, 328 model representation, 18 model selection, 364, 370–377 multiclass, 430–432, 437 in multimedia data mining, 596 neural networks for, 19, 398–408 pattern-based, 282, 318 perception-based, 348–350 precision measure, 368–369 as prediction problem, 328 process, 328 process illustration, 329 random forests, 382–383 recall measure, 368–369 robustness, 369 rough set approach, 427–428, 437 Yüklə 7,95 Mb. Dostları ilə paylaş:
|