Data Mining for Association Rules and Sequential Patterns: Sequential and Parallel Algorithms

ISBN-13: 9781461265115 / Angielski / Miękka / 2012 / 254 str.

Jean-Marc Adamo

Opis

Data mining includes a wide range of activities such as classification, clustering, similarity analysis, summarization, association rule and sequential pattern discovery, and so forth. The book focuses on the last two previously listed activities. It provides a unified presentation of algorithms for association rule and sequential pattern discovery. For both mining problems, the presentation relies on the lattice structure of the search space. All algorithms are built as processes running on this structure. Proving their properties takes advantage of the mathematical properties of the structure. Part of the motivation for writing this book was postgraduate teaching. One of the main intentions was to make the book a suitable support for the clear exposition of problems and algorithms as well as a sound base for further discussion and investigation. Since the book only assumes elementary mathematical knowledge in the domains of lattices, combinatorial optimization, probability calculus, and statistics, it is fit for use by undergraduate students as well. The algorithms are described in a C-like pseudo programming language. The computations are shown in great detail. This makes the book also fit for use by implementers: computer scientists in many domains as well as industry engineers.

Spis treści

1. Introduction.- 2. Search Space Partition-Based Rule Mining.- 2.1 Problem Statement.- 2.1.1 Canonical Attribute Sequences (cas).- 2.1.2 Database.- 2.1.3 Support.- 2.1.4 Association Rule.- 2.1.5 Problem Statement.- 2.2 Search Space.- 2.3 Splitting Procedure.- 2.4 Enumerating ?-Frequent Attribute Sets (cass).- 2.5 Sequential Enumeration Procedure.- 2.6 Parallel Enumeration Procedure.- 2.6.1 Initial Load Balancing.- 2.6.2 Computing the Starting Sets.- 2.6.3 Enumeration Procedure.- 2.6.4 Dynamic Load Balancing.- 2.7 Generating the Association Rules.- 2.7.1 Sequential Generation.- 2.7.2 Parallel Generation.- 3. Apriori and Other Algorithms.- 3.1 Early Algorithms.- 3.1.1 AIS.- 3.1.2 SETM.- 3.2 The Apriori Algorithms.- 3.2.1 Apriori.- 3.2.2 AprioriTid.- 3.3 Direct Hashing and Pruning.- 3.3.1 Filtering Candidates.- 3.3.2 Database Trimming.- 3.3.3 The DHP Algorithm.- 3.4 Dynamic Set Counting.- 4. Mining for Rules over Attribute Taxonomies.- 4.1 Association Rules over Taxonomies.- 4.2 Problem Statement and Algorithms.- 4.3 Pruning Uninteresting Rules.- 4.3.1 Measure of Interest.- 4.3.2 Rule Pruning Algorithm.- 4.3.3 Attribute Presence-Based Pruning.- 5. Constraint-Based Rule Mining.- 5.1 Boolean Constraints.- 5.1.1 Syntax.- 5.1.2 Semantics.- 5.1.3 Propagation of Boolean Constraints.- 5.2 Prime Implicants.- 5.3 Problem Statement and Algorithms.- 6. Data Partition-Based Rule Mining.- 6.1 Data Partitioning.- 6.1.1 Building a Probabilistic Model.- 6.1.2 Bounding Large Deviations for One cas (Chernoff bounds).- 6.1.3 Bounding Large Deviations for Sets of cass.- 6.2 cas Enumeration with Partitioned Data.- 6.2.1 Data Partitioning.- 6.2.2 Local ?-Frequent cas Generation.- 6.2.3 Global ?-Frequent cas Generation.- 7. Mining for Rules with Categorical and Metric Attributes.- 7.1 Interval Systems and Quantitative Rules.- 7.2 k-Partial Completeness.- 7.3 Pruning Uninteresting Rules.- 7.3.1 Measure of Interest.- 7.3.2 Attribute Presence-Based Pruning.- 7.4 Enumeration Algorithms.- 8. Optimizing Rules with Quantitative Attributes.- 8.1 Solving 1-1-Type Rule Optimization Problems.- 8.1.1 Problem Statement.- 8.1.2 MC\S Problem.- 8.1.3 MS\C Problem.- 8.1.4 MG Problem.- 8.2 Solving d-1-Type Rule Optimization Problems.- 8.3 Solving 1-q-Type Rule Optimization Problems.- 8.3.1 Problem Statement.- 8.3.2 MS\C Problem.- 8.3.3 MG Problem.- 8.4 Solving d-q-Type Rule Optimization Problems.- 8.4.1 Problem Statement.- 8.4.2 Basic Enumeration.- 8.4.3 Enumeration with Pruning.- 8.4.4 Pruning the Instantiation Set.- 9. Beyond Support-Confidence Framework.- 9.1 A Criticism of the Support-Confidence Framework.- 9.2 Conviction.- 9.3 Pruning Conviction-Based Rules.- 9.3.1 Analyzing Conviction.- 9.3.2 Transitivity-Based Pruning.- 9.3.3 Improvement-Based Pruning.- 9.4 One-Step Association Rule Mining.- 9.4.1 Building a Procedure for One-Step Mining.- 9.4.2 Building a Procedure for Improvement-Based Pruning.- 9.5 Correlated Attribute-Set Mining.- 9.5.1 Collective Strength.- 9.5.2 Correlated Attribute-Set Enumeration.- 9.6 Refining Conviction: Association Rule Intensity.- 9.6.1 Measure Construction.- 9.6.2 Properties.- 9.6.3 Relating ?-int(s ? u) to conv(s ? u).- 9.6.4 Mining with the Intensity Measure.- 9.6.5 ?-Intensity Versus Intensity as Defined in [G96].- 10. Search Space Partition-Based Sequential Pattern Mining.- 10.1 Problem Statement.- 10.1.1 Sequences of cass.- 10.1.2 Database.- 10.1.3 Support.- 10.1.4 Problem Statement.- 10.2 Search Space.- 10.3 Splitting the Search Space.- 10.4 Splitting Procedure.- 10.5 Sequence Enumeration.- 10.5.1 Extending the Support Set Notion.- 10.5.2 Join Operations.- 10.5.3 Sequential Enumeration Procedure.- 10.5.4 Parallel Enumeration Procedure.- Appendix 1. Chernoff Bounds.- Appendix 2. Partitioning in Figure 10.5: Beyond 3rd Power.- Appendix 3. Partitioning in Figure 10.6: Beyond 3rd Power.- References.