Machine Learning and Its Applications: Advanced Lectures (Lecture Notes in Computer Science, 2049)

✍ Scribed by Georgios Paliouras (editor), Vangelis Karkaletsis (editor), Constantine D. Spyropoulos (editor)

• Publisher: Springer
• Year: 2001
• Tongue: English
• Leaves: 334
• Category: Library

✦ Synopsis

In recent years machine learning has made its way from artificial intelligence into areas of administration, commerce, and industry. Data mining is perhaps the most widely known demonstration of this migration, complemented by less publicized applications of machine learning like adaptive systems in industry, financial prediction, medical diagnosis and the construction of user profiles for Web browsers.
This book presents the capabilities of machine learning methods and ideas on how these methods could be used to solve real-world problems. The first ten chapters assess the current state of the art of machine learning, from symbolic concept learning and conceptual clustering to case-based reasoning, neural networks, and genetic algorithms. The second part introduces the reader to innovative applications of ML techniques in fields such as data mining, knowledge discovery, human language technology, user modeling, data analysis, discovery science, agent technology, finance, etc.

✦ Table of Contents

Machine Learning and Its Applications
Preface
Table of Contents
Comparing Machine Learning and Knowledge Discovery in DataBases: An Application to Knowledge Discovery in Texts
Introduction
Reasoning by Induction
Hempel’s Paradox and the Theory of Confirmation
Implications that Carry the Meaning of Causality
Practical Consequences
What Makes Knowledge Discovery in Data (KDD) Different
“Previously Unknown Knowledge”
“Useful Knowledge”
Merging Approaches
Accurate vs. Understandable (and Useful) Knowledge
An Epistemological Difference
Engineering Is Not Dirty
Conclusion
Knowledge Discovery in Texts (KDT) Defined as a KDD
Conclusion
References
Learning Patterns in Noisy Data: The AQ Approach
Introduction
Multicriterion Selection of the Best Description
Completeness, Consistency, and Consistency Gain
A Definition of Description Quality
Empirical Comparison of Description Quality Measures
Admitting Inconsistency in AQ
Star Generation
Star Termination
Unexpected Difficulty
The Effect of Q(w) on Generated Rules
Admitting Incompleteness in AQ
Summary
References
Unsupervised Learning of Probabilistic Concept Hierarchies
Introduction
Incremental Formation of Probabilistic Concept Hierarchies
Representation and Organization of Knowledge
Performance and Learning Mechanisms
Empirical Studies of {sc Cobweb}
Basic Experimental Results
Effect of the Evaluation Function
The Effect of Search Control
The Effect of Missing Information
Extensions to C{relax fontsize {7}{8}selectfont bf OBWEB}
Minimizing Effects of Noise and Training Order
Learning Overlapping Concepts
Concept Formation in Temporal Domains
Other Variations on the C{relax fontsize {7}{8}selectfont bf OBWEB} Framework
Relation to Other Research on Unsupervised Learning
The A{relax fontsize {7}{8}selectfont bf UTO}C{relax fontsize {7}{8}selectfont bf LASS} Family
Induction of Bayesian Networks
Closing Remarks
References
Function Decomposition in Machine Learning
Introduction
Review of Function Decomposition Related to Machine Learning
Learning by Function Decomposition in {sf HINT}
Basic Decomposition Step
The Method
Some Properties of Basic Decomposition Step
Efficient Derivation of Incompatibility Graph
Partition Selection Measures
Overall Function Decomposition
Decomposition Algorithm
Complexity of Decomposition Algorithm
Attribute Redundancy and Decomposition-Based Attribute Subset Selection
Experimental Evaluation
Data Sets
Generalization
Hierarchical Concept Structures
Conclusion
References
How to Upgrade Propositional Learners to First Order Logic: A Case Study
Introduction
Knowledge Representation
Attribute Value Learning
First Order Representations
Background Knowledge
Note
The Propositional Learner {sc CN2}
Upgrading {sc CN2}
The Propositional Task and Algorithm
Examples Are Interpretations
First Order Hypotheses
Structuring the Search Space
Adapting the Search Operators
The Need for Bias
Implementing the Algorithm
Evaluation of ILP System
Extensions to the Basic System
Some Experimental Results with {sc ICL}
Experimental Settings
Propositional Data
Relational Data
Related Work and Conclusions
References
Case-Based Reasoning
1 Introduction
1.1 CBR Types and the CBR Cycle
2 Some Representative CBR Systems
3 Case Based Reasoning as a Learning Paradigm
4 Components, Issues, and Problems
4.1 Indexing/Retrieval/Selection
4.2 Memory Organization
4.3 Adaptation/Evaluation
4.4 Forgetting
4.5 Integration with Other Techniques
4.6 Uncertainty, Imprecision, and Incompleteness in CBR
5 Fuzzy CBR: Value Added to Conventional CBR
6 A Brief Account of Fuzzy Case-Based Reasoning Systems
6.1 The ARC System
6.2 The BOLERO System
6.3 The CAREFUL System
6.4 The CARS System
6.5 The FLORAN System
7 Conclusions
References
Genetic Algorithms in Machine Learning
Introduction
What Is a Genetic Algorithm
Genetic Algorithms in Machine Learning
History of Evolutionary Computing
The Basics of Genetic Algorithms
Elements of a Simple Genetic Algorithm
A Simple Genetic Algorithm
Extensions to the Basic Genetic Algorithm
Examples of Genetic Algorithms in Machine Learning
Learning Rule Sets
Evolving Computer Programs
Genetic Algorithms and Neural Networks
Genetic Algorithms and Reinforcement Learning
Interaction of Evolution and Learning
Combining Genetic Algorithms with Local Search Algorithms
The Baldwin Effect
Conclusions
References
Pattern Recognition and Neural Networks
Introduction
Bayes Decision Theory
Discriminant Functions and Discriminant Surfaces
Estimation of Unknown Probability Density Functions
Nearest Neighbor Rules
Linear Classifiers
Neural Network Classifiers
Choosing the Size of the Network
The Back Propagation Algorithm
Training Aspects
Radial Basis Function (RBF) Networks
Concluding Remarks
References
Model Class Selection and Construction: Beyond the Procrustean Approach to Machine Learning Applications
Introduction
Selection of a Machine Learning Method
Learning Methods
The Role of Model Class Selection in Learning
Methods for Model Class Selection
Data Transformation and Construction of New Learning
Conclusions and Further Issues
References
Integrated Architectures for Machine Learning
1 Introduction
2 Type of Interaction
3 Level of Interaction
4 Conclusions
References
The Computational Support of Scientific Discovery
Introduction
Stages of the Discovery Process
The Developer's Role in Computational Discovery
Some Computer-Aided Scientific Discoveries
Stellar Taxonomies from Infrared Spectra
Qualitative Factors in Carcinogenesis
Chemical Predictors of Mutagens
Quantitative Laws of Metallic Behavior
Quantitative Conjectures in Graph Theory
Temporal Laws of Ecological Behavior
Structural Models of Organic Molecules
Reaction Pathways in Catalytic Chemistry
Other Computational Aids for Scientific Research
An Illustration of Interactive Discovery
Progress and Prospects
References
Support Vector Machines: Theory and Applications
Introduction
A Brief Overview of the SVM Theory
Statistical Learning Theory: A Primer
Support Vector Machines Formulation
SVM Implementation
Experiments with SVM and Some Variations
Application of SVM to Medical Decision Support
Time Series Prediction Using Local SVM
An Application of SVM to Face Authentication
Conclusions
References
Pre- and Post-processing in Machine Learning and Data Mining
1 Introduction: Knowledge Discovery and Machine Learning
2 Description of Workshop Papers
Collecting and Preprocessing of Data
Postprocessing
3 Conclusion
References
Machine Learning in Human Language Technology
Introduction
Brief History
The ACAI ’99 Workshop on ML in HLT
Overview of the Papers
Conclusion
References
Machine Learning for Intelligent Information Access
1 Introduction
2 Overview of the Workshop Papers
3 Discussion
References
Machine Learning and Intelligent Agents
1 Introduction
2 Overview of the Workshop Papers
2.1 Intelligent Agents for the Web
2.2 Learning Interface Agents
2.3 Learning to Build Robot Maps
2.4 Learning to Take Decisions in a Real-Time Crisis
2.5 Inductive Concept Learning in a Three-Valued Logical Setting
3 Conclusions
References
Machine Learning in User Modeling
1 Introduction
2 Machine Learning in User Modeling
3 Overview of the Workshop Papers
3.1 Adaptive User Interfaces
3.2 Intelligent Agents and Multi-agent Systems
3.3 Student Modeling
3.4 Web Mining
4 Discussion
References
Data Mining in Economics, Finance, and Marketing
1 Introduction
2 Overview of the Workshop Papers
3 Conclusions
References
Machine Learning in Medical Applications
1 Introduction
3 Overview of the Workshop Papers
4 Discussion
References
Machine Learning Applications to Power Systems
1 Introduction
2 Description of Workshop Papers
2.1 Machine Learning Applications at the Power System Level
2.2 Machine Learning Applications at the Power System Component Level
3 Conclusions
References
Intelligent Techniques for Spatio-Temporal Data Analysis in Environmental Applications
1 Introduction
2 Short Description of Workshop Papers
3 Conclusions
References
Author Index