Intelligent Data Engineering and Automated Learning - IDEAL 2002: Third International Conference, Manchester, UK, August 12-14 Proceedings (Lecture Notes in Computer Science, 2412)

Intelligent Data Engineering and Automated Learning - IDEAL 2002
Preface
Organization
Table of Contents
Mining Frequent Sequential Patterns under a Similarity Constraint
	Introduction
	Basic Notions
	Similarity Constraint
	Experiments
	Conclusion
	References
Pre-pruning Classification Trees to Reduce Overfitting in Noisy Domains
	Introduction
	Overfitting of Classification Rules to Data
	Using the J-measure for Pre-pruning Classification Trees
	Experiments with Noisy Datasets
	Conclusions
	References
Data Mining for Fuzzy Decision Tree Structure with a Genetic Program
	1 Introduction
	A Brief Introduction to Fuzzy Sets, Fuzzy Logic, and the Fuzzy RM
	Data Mining with Genetic Programs
	Genetic Programs
	Evolving the Fuzzy Kinematic-ID Subtree
		Data Base Construction
		Terminal and Function Sets
		Genetic Program Evolved IPDT Subtrees
	Summary
	References
Co-evolutionary Data Mining to Discover Rules for Fuzzy Resource Management
	Introduction
	A Brief Introduction to Fuzzy Sets, Fuzzy Logic, and the Fuzzy RM
	Optimization of the Root Concept’s Parameters Using a Genetic Algorithm for Data Mining
	Co-evolutionary Data Mining
		Tools for Visualization of Data Mined Information
		Criterion for Re-optimization
		Stopping Criterion for Co-evolution
		A Simple Example of Co-evolutionary Optimization Using the Fuzzy Concept \"close\"
	Improvements in the RM’s Response through Data Mining
	Summary
	References
Discovering Temporal Rules from Temporally Ordered Data
	Introduction
	Formal Representation of the Problem
	Method
	Concluding Remarks
	References
Automated Personalisation of Internet Users Using Self-Organising Maps
	Introduction
	Methods
		Data Sets
		SOM Analysis
	Results and Discussion
	Conclusion
	References
Data Abstractions for Numerical Attributes in Data Mining
	Introduction
	Related Works
	Preliminaries
	Mining Association Rules with Data Abstractions
		Selecting Condition Attributes
		Finding Appropriate Data Abstractions
		Extracting Association Rules
		Extracting Rules with Multiple Conditions
	Experimental Results
	Concluding Remarks
	References
Calculating Aggregates with Range-Encoded Bit-Sliced Index
	Introduction
		Bit-Sliced Index
		Bitmap Encoding Schemes
	Evaluating Range and Aggregates with Bit-Sliced Index
	The Limitations of the Previous Techniques
	Calculating Aggregates with Range-Encoded Bit-Sliced Index
	Conclusions
	References
T3: A Classification Algorithm for Data Mining
	Introduction
	Description of C4.5 and T2
	Comparing T2 with C4.5
	T3: An Enhancement of T2
	Experimental Results
	Performance Evaluation
	Conclusions and Future Work
	References
A Hierarchical Model to Support Kansei Mining Process
	Introduction
	Requirements for Kansei User Model Adaptation
	Hierarchical Meta Schemas to Support Multimedia Data Mining Processes
	Conclusion
	References
Evolving SQL Queries for Data Mining
	Introduction
	Evolving SQL Queries
		Evolutionary Search
		Discussions
	Experimental Studies
		The Zoo Dataset
		Monk\'s Problems
		Credit Card Approval
		Discussion of the Results
	Conclusions
	References
Indexing and Mining of the Local Patterns in Sequence Database
	Introduction
	Problem Definition
	Method for Discovering LSPs
	Experimental Results
	Conclusions
	References
A Knowledge Discovery by Fuzzy Rule Based Hopfield Network
	Introduction
	Fuzzy Rule Based Hopfield Network
		Learning for Determining the Parameters of Gaussian Membership Functions
		Fuzzy Hopfield Network
	Proposed Rule Extraction Algorithm
	Experimental Results
	Conclusions
	References
Fusing Partially Inconsistent Expert and Learnt Knowledge in Uncertain Hierarchies
	Introduction
	Learnt and Expert Knowledge
		An Uncertain Hierarchical Representation of Knowledge
	Fusion in Uncertain Hierarchies
	Default Reasoning for Inconsistency Resolution
		Default Uncertain Inheritance Algorithm
		Practical Reasoning
	Conclusions
	References
Organisational Information Management and Knowledge Discovery in Email within Mailing Lists
	Introduction
	Document Analysis and Management: The Case of Email
	The Sentinel System
	Future Work
	Conclusion
	References
Design of Multi-drilling Gear Machines by Knowledge Processing and Machine Simulation
	Introduction
	Automated Configuration of Multi-drilling Gears
		Board Data and Machine Data
		Pattern Identification and Storage
		Initial Number of Supports, Gears, and Spindles
		Generalised Pre-placement and Computation of Board Complexity
		Iterative Configuration Process for Multi-drilling Gears
	Conclusion
	References
Classification of Email Queries by Topic: Approach Based on Hierarchically Structured Subject Domain
	Introduction
	Hierarchical Domain Structure
	Regular Expressions and Classifier\'s Dictionary
	Classification Rules
	Implementation of the Classifying Rules
	Used Tools
	Future Work
	References
A Knowledge-Based Information Extraction System for Semi-structured Labeled Documents
	Introduction
	Domain Knowledge Specification by XML
	Knowledge-Based Wrapper Induction
	Evaluation
	Conclusion
	References
Measuring Semantic Similarity Between Words Using Lexical Knowledge and Neural Networks
	Introduction
	Methodology
		Semantic Knowledge from WordNet
		Computing Semantic Similarity Using Neural Networks
	Experiment
		Deriving Similarity from Neural Networks
		Discussion
	Conclusion
	References
Extraction of Hidden Semantics from Web Pages
	Introduction
	Modeling Static Web Pages
	Dynamic Pages Issues
	Conclusions and Future Work
	References
Self-Organising Maps for Hierarchical Tree View Document Clustering Using Contextual Information
	Introduction
	Document Clustering Overview
		Indexing Methods
		Feature Selection and Term Weightings
		Clustering
	Related Work
	The Proposed Method
		Document Pre-processing
		SOM Procedure
	Experimentation and Discussion
	Conclusion and Future Work
	References
Schema Discovery of the Semi-structured and Hierarchical Data
	Introduction
	Concepts and Definitions
	Schema Discovery
		Constructing Set D and T from OEM Graph
		Algorithm for Schema Discovery
	Transforming the Extracted Schema to the Schema Tree
	Experiment
	References
RSTIndex: Indexing and Retrieving Web Document Using Computational and Linguistic Techniques
	Introductions and Previous Work
	Rhetorical Structure Theory
	The RSTIndex System
		Keyword Extraction
		Capturing the Document Linguistic Structure
		RST in Document Indexing
		Capturing the Documents Theme
	Conclusions
	References
A Case-Based Recognition of Semantic Structures in HTML Documents
	Introduction
	A Case-Based Transformation with Recognizing Semantic Structures
	Alignment for Identifying Semantic/Logical Structures
	Experimental Evaluation
	References
Expeditious XML Processing
	Introduction
	CXMLParser
		The Representation
		The Operations
	Performance Testing and Comparison
	Future Work and Conclusions
	References
Document Clustering Using the 1 + 1 Dimensional Self-Organising Map
	Introduction
	Document Clustering
	Related Work
		Introduction to the SOM
		WEBSOM
		The Growing Hierarchical SOM
	Proposed Methods and Results
	Conclusion
	References
Natural Language Processing for Expertise Modelling in E-mail Communication
	Introduction
	Related Work
	Descriptions of EMNLP
	Experimental Results
	Future Work
	References
A Branch and Bound Algorithm for Minimum Cost Network Flow Problem
	Introduction
	The Branch and Bound Algorithm
		Pruning the Searching Tree
	Computational Experiences
	Conclusion
	References
Study of the Regularity of the Users\' Internet Accesses
	Introduction
	Context of the Experimentation
	Description of the Regularities Analysis Method
	Batch Analysis
	Temporal Analysis
	Comparative Analysis of Users
	Related Works
	Conclusion
	References
An Intelligent Mobile Commerce System with Dynamic Contents Builder and Mobile Products Browser
	Introduction
	Related Work
	Proposed System
		Basic Designs Concepts
		Systems Architecture
	Implementation and Evaluation
		System Implementation
		System Evaluation
	Conclusions
	Reference
Focused Crawling Using Fictitious Play
	Introduction
	Coordination Model
		Basic Definitions
		Fictitious Play
		Connections to the Bayesian Inference
		Fictitious Play as a Coordination Method
	Test Settings
		Test Graphs
		Other Compared Methods
	Results
	Conclusions
	References
A User Adaptive Mobile Commerce System with a Middlet Application
	Introduction
	Proposed System
		Basic Designs Concepts
		Systems Architecture
	Implementation and Evaluation
		Systems Implementation
		System Evaluation
	Conclusion
	References
Weight-Vector Based Approach for Product Recommendation in E-commerce
	Introduction
	System Details
		Calculation of Similarities
	Weight Vectors and Learning
		Learning
		Learning Factor and Weight Adjustment
	Trend Change: Detection and Adjustment
	Comparisons and Conclusions
	References
The Development of an XML-Based Data Warehouse System
	Introduction
	Related work
		Traditional Client/Server Data Warehouse
		Web-Based Data Warehouse
		Data Cube and Star Schema
	The XML-Based Data Warehouse System Architecture
		The System Architecture
		The Methodology to Develop an XML Data Cube
	The Prototype System and System Evaluation
	Conclusions and Future Development
	Reference
Identifying Data Sources for Data Warehouses
	Introduction and Motivation
	Data Warehouse Source Identification Process
	The rmSeq Inference Algorithm
	A Simple Dependency Analysis Example
	Interpreting the mSeq Results with Dependency Graphs
	Conclusion
	References
Coordinating Learning Agents via Utility Assignment
	Introduction
	Utility Assignment
	Marketplace Application
		Agent Behaviour
		Utility Functions
		Results and Observations
		References
AGILE: An Agent-Assisted Infrastructure to Support Learning Environments
	Introduction
	Educational Resources
	Users and Roles
	Learning Agents
	Experimental Implementation
	Conclusions
	References
Multi-agent Fuzzy Logic Resource Manager
	Introduction
	Fuzzy Logic, Genetic Algorithms, Genetic Programs and the RM
	Subtrees of the RM
		The Isolated Platform Decision Tree
		The Multi-platform Decision Tree
		The Fuzzy EA Decision Algorithm
		The Fuzzy Parameter Selection Tree
		The Fuzzy Strategy Tree
	Validation of the RM Expert System
	Summary
	References
Transactional Multiple Agents
	Introduction
	Multi Transactional Agents
	The Scheme
	Conclusions
	References
An Information Model for a Merchant Trust Agent in Electronic Commerce
	Introduction
	Related Work
	The Proposed Model
		Existence
		Affiliation
		Policy
		Fulfillment
	The Model
	Conclusion
	References
MASIVE: A Case Study in Multiagent Systems
	Introduction
	IETAL Preliminaries
	Intrinsic Representations
	MASIVE: The World of Petitagés
		Social Hierarchies
		The Protolanguage in MASIVE
	Conclusions and Further Work
	References
Learning Multi-agent Strategies in Multi-stage Collaborative Games
	Introduction
	Sequential Reinforcement Learning for Multiple Agent
	Results
	Conclusions
	References
Emergent Specialization in Swarm Systems
	Introduction
	Learning Algorithm
	Experimental Results
		Homogeneous Teams
		Heterogeneous Teams
	Conclusions
	References
Distributed Mobile Communication Base Station Diagnosis and Monitoring Using Multi-agents
	Introduction
	The Need for a Multi-agent System
		Power Supply Modules and Environmental Factors
		Why Should a Multi-agent System Be Used?
	The Multi-agent System Infrastructure
		The Base Station Monitoring Agent
		The Mediation Agent
		The Personal Assistant Agent
	Design of Multi-agent System
	Conclusion and Future Work
	References
ABBA - Agent Based Beaver Application - Busy Beaver in Swarm
	Introduction
	The Problem
	The Model
	The Environment
	Learning Techniques
	The Simulaton
	Results
	Future Works
	References
Centralised and Distributed Organisational Control
	Background and Motivation
	Initial Concept Demonstrator Overview
	Simulation Scenarios and Results
	Discussion and Future Work
	References
Mining Dependence Structures from Statistical Learning Perspective
	Basic Issues of Statistical Learning
	Dependence among Samples from a One-Object World
	Dependence among Samples from a Multi-object World
		Dependence among Samples from a Multi-object World
		Mining Dependence Structure across Invisible Multi-objects
	A Key Challenge and Existing Solutions
		A Key Challenge and Existing Solutions
		Efforts in the First Stream
		Efforts in the Second Stream
	Bayesian Ying-Yang Harmony Learning
	References
k-Means - A Generalized -Means Clustering Algorithm with Unknown Cluster Number
	Introduction
	A Metric for Data Clustering
	Qualitative Analysis of MAP-Based Data Clustering Assignment
	$k^*$-Means Algorithm
	Experimental Results
	Conclusion
	References
Multiagent SAT (MASSAT): Autonomous Pattern Search in Constrained Domains
	Introduction
		Satisfiability Problems
		Multiagent Systems
	The MASSAT Approach
		An Environment
		An Agent
		The System Schedule
	Experimental Results
	Conclusion and Future Work
	References
A Text Mining Agents Based Architecture for Personal E-mail Filtering and Management
	Introduction
	System Architecture
	USPC Agent
	R2L Agent
		RS-ILP
		Learning Process in R2L
	Related Work
	Conclusions
	References
Framework of a Multi-agent KDD System
	Introduction
	Architecture of the GLS System
	Ontology of KDD Agents
	Planning Mate Agent (PMA)
	Controlling Mate Agent (CMA) and KDD Agents
	Conclusions
	References
Intraday FX Trading: An Evolutionary Reinforcement Learning Approach
	Introduction
	Literature Review
	The Problem Defined
		Modelling Trading
		Technical Indicators
		Trading Strategies
		Evaluation
	Applying RL to the Technical Trading Problem
	Evolutionary Reinforcement Learning Approach
	Numerical Experiments
	Discussion and Future Work
	References
An Up-Trend Detection Using an Auto-Associative Neural Network: KOSPI 200 Futures
	Introduction
	Definition of Up-Trend
	Auto-Associative Neural Network as an Up-Trend Detector
	Data Collection and Neural Network Training
	Results
	Conclusions
	References
Stock Price and Index Forecasting by Arbitrage Pricing Theory-Based Gaussian TFA Learning
	Introduction
	Review on Arbitrage Pricing Theory
	Overview of Temporal Factor Analysis
	Using Gaussian TFA for Stock Index Prediction
		Data Considerations
		Data Preprocessing
		Experimental Results
		Performance Evaluation
	Conclusion
	References
A Comparative Study on Three MAP Factor Estimate Approaches for NFA
	Introduction
	Approaches for MAP Estimate Problem in NFA
		Iterative FPA
		Gradient Descent Approach
		Conjugate Gradient Method
		Gaussian Approximation Initialization
	Experimental Demonstration of NFA
		Data Description
		Experimental Results
	Discussion on the MAP Factor Estimate Process
		A Typical MAP Factor Estimate Process
		Discussion on Estimate Accuracy
		Discussion on Estimate Efficiency
		Concluding Remarks
		References
A Neural Classifier with Fraud Density Map for Effective Credit Card Fraud Detection
	Introduction
	Fraud Detection Scheme with Fraud Density Map
		Fraud Detection System
		Fraud Density Map
		Combining the Fraud Score and the Fraud Density
	Experimental Results
	Conclusions
	References
A Comparison of Two Techniques for Next-Day Electricity Price Forecasting
	Introduction
	Structure of the Spanish Electricity Market
	k Weighted Nearest Neighbours
		Test Results
	Dynamic Regression
		Results
	Conclusions
	References
Support Vector Machine Regression for Volatile Stock Market Prediction
	Introduction
	Support Vector Regression
	Experiments
	Discussion and Conclusion
	References
Complexity Pursuit for Financial Prediction
	Complexity Pursuit
	Logarithmic Differences
	Comparing the Methods on Real Data
		Results
	Conclusion
	References
Artificial Intelligence in Portfolio Management
	Introduction
	Artificial Intelligence Technologies
		Genetic Algorithm
		Fuzzy Network
		Fuzzy Network Synthesized with Genetic Algorithm
	Intelligent Portfolio Management System
		Stock Rating Subsystem
		Asset Allocation Optimization Subsystem
	Experiments
	Conclusion and Discussion
	References
The Multilevel Classification Problem and a Monotonicity Hint
	Introduction
	Problem Setup
		Incorporating the Monotonicity Hint
	Experimental Simulations and Further Study
	References
Adaptive Filtering for GARCH Models
	Introduction
	Volatility Definitions
	The GARCH(1,1) Process
		GARCH(1,1) Volatility Forecast
	Adaptive Filtering
		Adaptive LMS Filter
		Applying $mathaccent \"705F {sigma }[tau ]$
	Results
	Conclusion
	References
Application of Self-Organising Maps in Automated Chemical Shift Correction of In Vivo H MR Spectra
	Introduction
	Methods
		In vivo 1H NMR Spectroscopy
		SOM Analysis
	Results and Discussion
	Conclusion
	References
Supervised Learning of Term Similarities
	Introduction
	Term Similarity Measure
		Contextual Similarity
		Functional Similarity
		Lexical Similarity
	Tuning a Similarity Measure with a Genetic Algorithm
	Experiments and Evaluation
	Conclusion
	References
BIKMAS: A Knowledge Engineering System for Bioinformatics
	Introduction
	Methodology
	Results
	Conclusions
	References
Unsupervised Feature Extraction of in vivo Magnetic Resonance Spectra of Brain Tumours Using Independent Component Analysis
	Introduction
		MR Data and Independent Component Analysis
		Preliminary Work with ICA on unhbox voidb @x hbox {$rm ^{1}H, $}MRS
	Data and Methods
	Results
	Discussion
	References
Fuzzy Rule-Based Framework for Medical Record Validation
	Introduction
	Motivating Examples
	FuzzyKlean Framework
	Experiments
	Conclusion
	References
Classification Learning by Decomposition of Numerical Datasets
	Introduction
	Our Decomposition Method
		Partition of the Attribute Set
		Discretization of Numerical Attributes
		Construction of Intermediate Decision Tables
		Determination of the Final Decision Table
	Experimental Results
	Conclusions
	References
Combining Feature Selection with Feature Weighting for k-NN Classifier
	Introduction
	Computation of Information Gain
	Feature Selection Based on CORE, Binary Mutual Information, and Class Mutual Information (FS-CBC)
	Experiments and Results
	Conclusions
	References
Pattern Selection for Support Vector Classifiers
	Introduction
	Proposed Algorithm for Pattern Selection
	Results
	Conclusion
	References
Graphical Features Selection Method
	Introduction
	Feature Selection and Classification
		Thrombin Data Set
		Numerical Results
	Shannon\'s Entropy
	Conclusion
	References
Fuzzy-Neural Inference in Decision Trees
	Introduction
	Generating A Fuzzy Decision Tree
		Fuzzification
		Pure Fuzzy Inference
	Fuzzy Neural Inference
	Experiments
		Experiments
		Results
	Conclusion
	References
Decision Tree Based Clustering
	Introduction
	Log Likelihood and Cross Entropy
		Approximated Log Likelihood
		Cross Entropy
	Bhattacharyya Distance Based Measures
		Bhattacharyya Distance
		Scatter Matrix
		Class Separability Using Bhattacharyya Distance
	Experiment
	Conclusions
	References
Usage of New Information Estimations for Induction of Fuzzy Decision Trees
	Introduction
	Preliminaries of Fuzzy Logic
	Information-Theoretic Learning
		Information and Summary Entropies in Fuzzy Sets
		Mutual Summary Information and Entropy for Sequence of Attributes
	The Algorithm of FDT Induction
	Conclusion
	References
Genetic Algorithm Based-On the Quantum Probability Representation
	Introduction
	GAQPR
		Representation and Updating of Chromosomes
		Procedure of GAQPR
		Crossover Operator
		Mutation Operator
	Experiments and Discussion
	Conclusion
	References
A Dynamic Method for Discretization of Continuous Attributes
	Introduction
	GroupMerge
	Empirical Evaluation
	Conclusion
	References
A New Neural Implementation of Exploratory Projection Pursuit
	Introduction
	Maximum Likelihood Hebbian Learning
	Results Using Artificial Data Sets
	Minimum Likelihood Hebbian Learning
	Experiments Using a Real Data Set
	Conclusions
	References
A General Framework for a Principled Hierarchical Visualization of Multivariate Data
	Introduction
	Probabilistic Model of Hierarchic Visualization
		Latent Trait Models
		Hierarchical Latent Trait Models
	Local Geometric Analysis of the Latent Trait Manifolds
	Hierarchical Visualization of Text Documents
	Conclusions
	References
Chinese Character Recognition - Comparison of Classification Methodologies
	Introduction
	Dominant Point Algorithm
		Digitization and Preprocessing
		Feature Extraction
	Classifier Systems
		Discriminant Analysis
		Machine Learning C4.5
		Fuzzy Nearest Neighbour Method
	Conclusions
	References
Lempel-Ziv Coding in Reinforcement Learning
	Introduction
	Source Coding
		Description as a Model of Information Source
		Lempel-Ziv Complexity
		Incremental Parsing of String
	Experiments
	Discussions and Conclusions
	References
Efficient Face Extraction Using Skin-Color Model and a Neural Network
	Introduction
	Image Segmentation
	Facial Region Extraction
	Experimental Results
	Conclusions
	References
Feature Weights Determining of Pattern Classification by Using a Rough Genetic Algorithm with Fuzzy Similarity Measure
	Introduction
	Problem Description
		Qualitative Similarity Information and Condition
		Fuzzy Similarity Measure
	Rough Genetic Algorithm for QSI
		Rough Genetic Algorithm (RGA)
		Simulation and Test I
	Application of Weight Vector
		Weighted Discretized Value Difference Metric
		Simulation and Test II
	Concluding Remarks
	References
Recursive Form of the Discrete Fourier Transform for Two-Dimensional Signals
	Introduction
	Recursive Procedures of the Fourier Transform
		Recursive Procedure for One-Dimensional Signals
		Recursive Procedure for Two-Dimensional Signals
	Computer Simulations
	Conclusion
	References
Viseme Recognition Experiment Using Context Dependent Hidden Markov Models
	Introduction
	Viseme Recognition
	Experimental Results
	Conclusions
	References
Stave Extraction for Printed Music Scores
	Introduction
	Stage-1: Extraction of Candidate Points
	Stage-2: Connection of Candidate Points Using DP Matching
	Stage-3: Composition of Stave Groups Using Labeling
	Stage-4: Extraction and Adjustment of the Edges of Staff Lines
	Experimental Results and Discussion
	Conclusions
	References
Scaling-Up Model-Based Clustering Algorithm by Working on Clustering Features
	Introduction
	Adaptive Data Summarization Procedure
	An EM Algorithm for Clustering Features
	Experiments
	Conclusion
	References
A New Approach to Hierarchically Retrieve MPEG Video
	Introduction
	Our Retrieval Principle
		Overview of Our Retrieval Approach
		Quickly Filter by Analyzing dct_dc_size Field
		Get Final Results by Analyzing DC Image
	Experimentation
	Conclusion
	References
Alpha-Beta Search Revisited
	Introduction
	Game-Tree Search
	Search Enhancements
	Performance
	Conclusions
	References
Quantifying Relevance of Input Features
	Introduction
	Quantifying Feature Relevance
	Measures of Ranking Performance
	Identifying Risk Factors of Osteoporosis
	Conclusion
	References
Author Index