Quality measures in pattern mining

For more info about our services contact : help@bestpfe.com

Table of contents

Chapter 1 Introduction
1.1 Outline of pattern mining
1.2 Thesis structure and contribution
Chapter 2 Background
2.1 Formal concept analysis
2.2 Pattern structures and interval pattern structures
2.3 Minimum description length principle
2.4 Quality measures in pattern mining
2.4.1 Pattern set quality measures in unsupervised settings
2.4.2 Quality measures in supervised settings
Chapter 3 Closure structure
3.1 Introduction
3.2 Related work
3.3 Basic notions
3.4 Closure structure and its properties
3.4.1 Keys and passkeys
3.4.2 Closure structure
3.4.3 Passkey-based order ideal
3.4.4 Assessing data complexity
3.4.5 Closeness under sampling
3.5 The GDPM algorithm
3.5.1 Computing the closure structure with GDPM
3.5.2 The extent-based version of GDPM
3.5.3 Complexity of GDPM
3.5.4 Related approaches to key-based enumeration
3.5.5 Computing passkeys. Towards polynomial complexity
3.6 Experiments
3.6.1 Characteristics of the datasets
3.6.2 Computational performance
3.6.3 Data topology or frequency distribution within levels
3.6.4 Coverage and overlaps
3.6.5 Usefulness of concepts
3.6.6 Case study
3.7 Discussion and conclusion
Chapter 4 Pattern mining in binary data
4.1 Introduction
4.1.1 Pattern types
4.1.2 Exploring the pattern search space
4.1.3 Interestingness measures
4.2 Minimum description length principle in itemset mining
4.3 Greedy strategy to reduce pattern search space
4.3.1 Likely-occurring itemsets
4.3.2 Likely-occurring itemsets and related notions
4.3.3 Experiments
4.4 Adapting the best practices of supervised learning to itemset mining
4.4.1 KeepItSimple: an algorithm for discovering useful pattern sets
4.4.2 Experiments
4.5 Discussion and conclusion
Chapter 5 Pattern mining in numerical data
5.1 Introduction
5.2 Related work
5.2.1 Numerical data preprocessing
5.2.2 MDL-based approaches to pattern mining
5.3 Basic notions
5.3.1 Formalization of data and patterns
5.3.2 Information theory and MDL
5.3.3 Derivation of the plug-in codes
5.3.4 ML-estimates of the parameters of the multinomial distribution
5.4 Mint
5.4.1 The model encoding
5.4.2 The Mint algorithm
5.5 Experiments
5.5.1 Datasets
5.6 Discussion and conclusion
6 Conclusion and future work
Bibliography