You are previewing Intelligent Soft Computation and Evolving Data Mining: Integrating Advanced Technologies.
O'Reilly logo
Intelligent Soft Computation and Evolving Data Mining: Integrating Advanced Technologies

Book Description

Intelligent Soft Computation and Evolving Data Mining: Integrating Advanced Technologies is a compendium that addresses this need. It integrates contrasting techniques of conventional hard computing and soft computing to exploit the tolerance for imprecision, uncertainty, partial truth, and approximation to achieve tractability, robustness and low-cost solution. This book provides a reference to researchers, practitioners, and students in both soft computing and data mining communities, forming a foundation for the development of the field.

Table of Contents

  1. Copyright
  2. Editorial Advisory Board
  3. List of Reviewers
  4. Foreword
  5. Preface
  6. 1. Integrating Soft Computation and Data Mining
    1. 1. Introduction to the Experimental Design in the Data Mining Tool KEEL
      1. ABSTRACT
      2. INTRODUCTION
      3. BACKGROUND
      4. KEEL
        1. Main Features
        2. Data Management
        3. Design of Experiments: Off-Line Module
        4. Computer-Based Education: On-Line Module
      5. EVOLUTIONARY ALGORITHMS IN KEEL
      6. EXPERIMENTAL EXAMPLES OF KEEL
        1. Classification Example
        2. Regression Example
        3. Unsupervised Learning Example
      7. FUTURE TRENDS
      8. CONCLUSION
      9. ACKNOWLEDGMENT
      10. REFERENCES
      11. ENDNOTES
    2. 2. Cat Swarm Optimization Supported Data Mining
      1. ABSTRACT
      2. INTRODUCTION
      3. CONCEPT OF DATA MINING
        1. Regression
        2. Classification
        3. Time-Series Forecasting
        4. Clustering
        5. Association Rule
        6. Sequence Discovery
        7. Definition of Interesting Rules and Patterns
      4. COMPUTATIONAL INTELLIGENCE AND EVOLUTIONARY COMPUTING
        1. Cat Swarm Optimization (CSO)
          1. The Seeking Mode Process
          2. The Tracing Mode Process
          3. Experimental Results
          4. Discussion
        2. Parallel Cat Swarm Optimization (PCSO)
          1. Parallel Tracing Mode Process
          2. Information Exchanging
          3. Experimental Results
          4. Discussion
      5. FEASIBLE SOLUTIONS OF COMPUTATIONAL INTELLIGENCE FOR DATA MINING
        1. Existing Criteria of Evaluation in Data Mining
        2. Feasible Solutions Combines Intelligence Soft Computing with Data Mining
      6. CONCLUSION
      7. REFERENCES
    3. 3. Dynamic Discovery of Fuzzy Functional Dependencies Using Partitions
      1. ABSTRACT
      2. INTRODUCTION
      3. A SIMILARITY-BASED FUZZY RELATIONAL DATA MODEL
        1. Similarity Relation and Domain Partition
        2. A Similarity-Based Fuzzy Relational Data Model
      4. FUZZY FUNCTIONAL DEPENDENCIES ON THE SIMILARITY-BASED DATA MODEL
        1. Fuzzy Functional Dependencies
        2. Validation of Fuzzy Functional Dependencies
      5. INCREMENTAL SEARCHING OF FUZZY FUNCTIONAL DEPENDENCIES
        1. The Incremental Searching Algorithm for Fuzzy Functional Dependencies
        2. Procedure 1: Compute Domain Partition for Attribute A W.R.T. Level Value αA
        3. Procedure 2: Compute tuple Partition of R W.R.T. Attribute A and Level ValueαA
        4. Procedure 3: Compute the Refinement of Two Tuple Partitions
        5. Procedure 4: Find Least Specialization of a Contradicted Functional Dependency
        6. Procedure 5: Remove More Specialized Functional Dependencies
      6. ANALYSIS
      7. NUMERICAL EXAMPLES
      8. CONCLUSION
      9. REFERENCES
    4. 4. An Intelligent Data Mining System Through Integration of Electromagnetism-Like Mechanism And Fuzzy Neural Network
      1. ABSTRACT
      2. INTRODUCTION
      3. FUZZY NEURAL NETWORKS
        1. Operations of Fuzzy Numbers
        2. FNN Learning Algorithm
      4. A META-HEURISTIC ALGORITHM FOR GLOBAL OPTIMIZATION - THE ELECTROMAGNETISM-LIKE MECHANISM (EM)
        1. General Scheme
          1. Initialization
          2. Local Search
          3. Calculation of Total Force Vector
          4. Movement According to Total Force Vector
      5. THEORETICAL STUDY OF THE EM
      6. ELECTROMAGNETISM-LIKE MECHANISM BASED FUZZY NEURAL NETWORK (EMFNN)
      7. COMPUTATIONAL RESULTS
        1. Example One
        2. Example Two
        3. Sales Forecasting Case
          1. Experiment 1
          2. Experiment 2
        4. Summary
      8. CONCLUSION AND FUTURE RESEARCH DIRECTIONS
      9. REFERENCES
      10. KEY TERMS AND DEFINITIONS
        1. Glossary
  7. 2. Soft Computation
    1. 5. Computational Intelligence-Revisited
      1. ABSTRACT
      2. INTRODUCTION
      3. DEFINITIONS OF CI
        1. Background
        2. Definition and Description
      4. CLASSIFICATION METHODS
        1. Different Classification Method
          1. Computational-Medium-Based Classification Method
          2. Parallelism Based Classification Method
          3. Natural-Biology-Based Classification Method
          4. Problem-Space-Based Classification Method
          5. Computational-Intelligent-Based Classification Method
        2. SMB Classification Method
      5. ORGANIC MECHANISM SIMULATION CLASS
        1. Group Mechanism Simulation Class
          1. Group Evolution Mechanism Simulation Class
          2. Group Collaboration Mechanism Simulation Class
          3. Nonlinear Mapping Model for Group Mechanism Simulation class
        2. Individual Mechanism Simulation Class
          1. Fuzzy Logic
          2. Neural Networks
          3. Nonlinear Mapping Model for Individual Mechanism Simulation class
      6. INORGANIC MECHANISM SIMULATION CLASS
      7. ARTIFICIAL MECHANISM SIMULATION CLASS
      8. FUTURE TRENDS
      9. CONCLUSION
      10. REFERENCES
    2. 6. Artificial Clonal Selection Model and Its Application
      1. ABSTRACT
      2. INTRODUCTION
        1. Background
          1. Natural Immune System
            1. Overview of the Immune System
            2. Immune Organs and Cells
            3. Immune Response
          2. Artificial Immune Systems (AIS)
            1. Negative Selection Based Algorithm
            2. Immune Network Theory Based Algorithm
            3. Clonal Selection Based Algorithm
        2. Advanced Clonal Selection Model
          1. Receptor Editing Operator
          2. Chaotic Initialization and Distance-Based Hypermutation
            1. Chaotic Initialization
            2. Distance-Based Hypermutation Operator
          3. Polyclonal Selection Algorithm with Simulated Annealing (PCASA)
            1. Greedy Crossover Operator
            2. Simulated Annealing Strategy
          4. Lateral Interactive Receptor Editing (LIRE)
            1. Fast Algorithm for the Implementation of LIRE
          5. CSA Combined with Ant Colony Optimization (CSACO)
            1. An Overview of Ant Colony Optimization
            2. Pheromone-Linker to Combine CSA with ACO
        3. Application to the Traveling Salesman Problem (TSP)
      3. FUTURE TRENDS
      4. CONCLUSION
      5. REFERENCES
    3. 7. Risk-Management Models Based on the Portfolio Theory Using Historical Data under Uncertainty
      1. ABSTRACT
      2. INTRODUCTION
      3. BACKGROUND
        1. Literature Review
        2. Notations of Parameters and Assumptions
        3. Mathematical Definitions of Fuzzy Set, Number and Variable
        4. Definition 1
        5. Definition 2
        6. Definition 3
        7. Definition 4
        8. Theorem 1
        9. Example 1
        10. Possibility and Necessity Measure
        11. Definition 5
        12. Definition 6
        13. Example 2
        14. Fuzzy Chance Constrained Programming
        15. Risk Management Model Based On Mean-Variance Theory
        16. Risk Management Model Based on Mean-Absolute Deviation Theory
        17. Risk Management Model Based on Mean-Variance Theory
      4. FUZZY EXTENSION OF RISK-MANAGEMENT MODELS
        1. Fuzzy Extension of Risk-Management Model with the Mean-Variance
        2. Fuzzy Extension of Risk-Management Model with the Mean-Absolute Deviation
        3. Fuzzy Extension of Risk-Management Models with the Safety-First Theory
        4. Solution Method
        5. Numerical Example
      5. FUTURE TRENDS
      6. CONCLUSION
      7. REFERENCES
    4. 8. Neuro-Fuzzy System Modeling
      1. ABSTRACT
      2. INTRODUCTION
      3. BACKGROUND
        1. Neural Networks
          1. Network Architectures
      4. LEARNING ALGORITHM
        1. Fuzzy Systems
          1. System Architecture
          2. Some Types of Fuzzy Systems
        2. Neuro-Fuzzy Systems
        3. Literature Survey
      5. NEURO-FUZZY SYSTEM MODELING
      6. STRUCTURE IDENTIFICATION
          1. Similarity-and-Merge-Based Rule Generation
          2. Data Partitioning Stage
        1. Cluster Merge Stage
        2. Fuzzy Rule Extraction
      7. PARAMETER IDENTIFICATION
          1. Zero-Order TSK-Type Fuzzy Neural Networks
          2. Hybrid Learning Algorithm
        1. Recursive SVD-Based Least Squares Estimator
        2. Theorem 3: Recursive SVD-Based Least Squares Estimator
        3. Proof:
        4. Gradient Descent Method
        5. First-Order TSK-Type Fuzzy Neural Networks
        6. An Example
      8. FUTURE TRENDS
      9. CONCLUSION
      10. REFERENCES
    5. 9. Network Based Fusion of Global and Local Information in Time Series Prediction with The Use of Soft-Computing Techniques
      1. ABSTRACT
      2. INTRODUCTION
      3. LOCAL PREDICTION APPROACHES
        1. Grey Model
        2. Fourier Grey Model (FGM)
        3. Markov Fourier Grey Model (MFGM)
      4. GLOBAL PREDICTION APPROACHES
        1. Backpropagation Neural Networks
        2. SONFIN
      5. STUDY OF FUSING APPROACHES FOR MULTI-STEP PREDICTION
        1. Neural Network Based Residual Correction Approach
        2. Global Based Neural Network Fusion
        3. Global Based SONFIN Fusion
        4. Kalman Filter Based Estimator
        5. Network Based Estimator
      6. CONCLUSION
      7. REFERENCES
    6. 10. Weights Direct Determination of Feedforward Neural Networks without Iterative BP-Training
      1. ABSTRACT
      2. INTRODUCTION
      3. BACKGROUND
      4. POWER-ACTIVATION NN WITH WEIGHTS DIRECTLY DETERMINED
        1. PANN Model Description & Theoretical Basis
        2. PANN Weights-Updating Formulas
        3. PANN Weights Direct Determination
        4. PANN Simulation & Verification
          1. Example 1 (Target Function Approximation)
          2. Example 2 (Nonlinear Dynamic System Identification)
            1. Training Procedure
            2. Testing Results
        5. PANN Section Summary
      5. LAGUERRE NN WITH WEIGHTS & STRUCTURAL DETERMINATION
        1. LNN Model Description
        2. LNN Theoretical Foundation
        3. LNN Weights Updating & Determination
        4. LNN Structure Automatic Determination
        5. LNN Approximation & Prediction Example
        6. LNN Section Summary
      6. GEGENBAUER NN WITH WEIGHTS & STRUCTURAL DETERMINATION
      7. COMPARISON AMONG PANN, LNN AND GNN
      8. FUTURE TRENDS
      9. CONCLUSION
      10. ACKNOWLEDGMENT
      11. REFERENCES
    7. 11. The Hopfield-Tank Neural Network for the Mobile Agent Planning Problem
      1. ABSTRACT
      2. INTRODUCTION
      3. BACKGROUND
        1. The Mobile Agent
        2. Hopfield Neural Network
        3. The Hopfield-Tank Neural Network
      4. THE MOBILE AGENT PLANNING PROBLEM
        1. The Objective Function
        2. Example
      5. THE MOBILE AGENT PLANNING MODEL
        1. The State Variables
        2. The Constraints and Problem Goal
          1. Constraint 1: The Starting and End Points of the Mobile Agent Must Be Site H
          2. Constraint 2: Visit At Most One Site In Each Time Slot
          3. Constraint 3: Visit The Same Site Exactly Once
          4. Constraint 4: Visit All The N Sites
          5. Constraint 5: The Legitimacy of a Tour
        3. Problem Goal: Minimum Completion Time for a Mobile Agent
        4. The MAP Energy Function
          1. Constraint 2: Visit At Most One Site In Each Time Slot
          2. Constraint 3: Visit The Same Site Exactly Once
          3. Constraint 4: Visit All The N Sites
          4. Constraint 5: The Legitimacy of a Tour
        5. Problem Goal: Minimum Completion Time for a Mobile Agent
        6. The Connection Weight Matrices
        7. Example
        8. The Activation Function
        9. The MAP Algorithm
      6. FUTURE TRENDS
      7. CONCLUSION
      8. REFERENCES
    8. 12. A Novel Neural Fuzzy Network Using a Hybrid Evolutionary Learning Algorithm
      1. ABSTRACT
      2. INTRODUCTION
      3. PARTICLE SWARM OPTIMIZATION, COOPERATIVE PARTICLE SWARM OPTIMIZATION AND CULTURAL ALGORITHM
        1. Particle Swarm Optimization
        2. Cooperative Particle Swarm Optimization
        3. Cultural Algorithm
      4. STRUCTURE OF FUNCTIONAL-LINK-BASED NEURAL FUZZY NETWORK
        1. Functional Link Neural Networks
        2. Structure of FLNFN Model
      5. THE CULTURAL COOPERATIVE PARTICLE SWARM OPTIMIZATION FOR THE FLNFN MODEL
        1. Experimental Results
          1. Example 1: Prediction of a Chaotic Signal
          2. Example 2: Prediction of Chaotic Time Series
          3. Example 3: Forecast of the Number of Sunspots
      6. CONCLUSION AND FUTURE WORKS
      7. REFERENCES
    9. 13. Power System Load Frequency Control Using Combined Intelligent Techniques
      1. ABSTRACT
      2. INTRODUCTION
      3. RELEVANT BACKGROUND AND CONVENTIONAL APPROACHES
        1. Brief Literature Review
        2. Electrical Power System Industry
        3. Conventional Load Frequency Controllers
      4. PHILOSOPHY OF INTELLIGENT CONTROLLER DESIGN
        1. Fuzzy Logic Controller
          1. FLC Design Procedure Overview
      5. GENETIC ALGORITHMS
        1. Artificial Neural Networks
          1. Multi-Layer Feed Forward Neural Networks
          2. Higher-Order Feed-Forward Networks
        2. Overall Load Frequency Intelligent Controller Scheme
          1. Stage 1: Fuzzy Part
          2. Stage 2: GA Part
          3. Stage 3: ANN Part
          4. Application and Results
      6. CONCLUSION
      7. REFERENCES
    10. A. APPENDIX
    11. 14. Computational Intelligence Clustering for Dynamic Video Watermarking
      1. ABSTRACT
      2. INTRODUCTION
      3. APPLYING COMPUTATIONAL INTELLIGENCE CLUSTERING METHODS TO VIDEO WATERMARKING
        1. K-Means Clustering
        2. Fuzzy C-Means Clustering and Thresholding
        3. Swarm Intelligence Clustering Algorithm
        4. Combining Swarm Intelligence and FCM Clustering
      4. PROPOSED VIDEO WATERMARKING SCHEME
        1. Watermark Embedding Procedure
        2. Watermark Extraction Procedure
      5. EXPERIMENTAL RESULTS
        1. Motion Vector Selection with Fixed Percentage
        2. Motion Vector Selection with K-means Clustering
        3. Motion Vector Selection with FCM Clustering
        4. Motion Vector Selection with Thresholding FCM
        5. Motion Vector Selection Using Swarm-Intelligence
        6. Motion Vector Selection with Swarm-Intelligence-Based FCM
      6. SUMMARY
      7. REFERENCES
    12. B. APPENDIX
      1. A. 1 FCM algorithm (Bezdek, 1973):
      2. A. 2 Swarm Intelligence Clustering algorithm (Bin, Yi, Shaohui, & Zhongzhi, 2002):
  8. 3. Data Mining
    1. 15. Data Mining Meets Internet and Web Performance
      1. ABSTRACT
      2. INTRODUCTION AND MOTIVATION
      3. NETWORK BACKGROUND
        1. Network Measurements
        2. Passive and Active Measurements
        3. Sound and Rational Measurements
        4. Internet Layer Performance
        5. Internet Data Sources
        6. TCP Throughput Prediction
      4. INTERNET AND WEB MINING BACKGROUND
        1. Data Sources
        2. Data Preparation
          1. Stage 1
          2. Stage 2
          3. Stage 3
          4. Stage 4
        3. Web Layer Performance
        4. Predictive Data Mining
      5. PROPOSED MODEL OF WEB TRANSACTION
      6. WING AND MWING MEASUREMENT SYSTEMS
      7. WING
      8. MWING
      9. REAL-LIFE EXPLORATORY CASES
        1. Case 1
          1. Experiment Setup
          2. Data Source and Data Preprocessing
          3. Mining Algorithms
          4. Results and Discussion
        2. Case 2
          1. Experiment Setup
          2. Data Source and Data Preprocessing
          3. Mining Algorithms
          4. Results and Discussion
      10. CONCLUSION
      11. ACKNOWLEDGMENT
      12. REFERENCES
    2. 16. Predicting Similarity of Web Services Using WordNet
      1. ABSTRACT
      2. INTRODUCTION
      3. SIMILARITY OF WEB SERVICES
        1. Lexical Similarity of Web Services
        2. Structure Similarity of Web Services
      4. DATA PRE-PROCESSING
        1. Stop Words Removal
        2. Stemming
      5. WORDNET BASED SEMANTIC SIMILARITY
        1. WordNet
        2. Organization of WordNet
        3. Word Sense Disambiguation
        4. Determine Word Sense Using WordNet
        5. Measure Words Similarity Using WordNet
        6. WordNet Based Similarity of Web Services
      6. CLUSTERING AND CLASSIFICATION OF WEB SERVICES
        1. Classification of Web Services
      7. EXPERIMENTAL RESULTS
      8. CONCLUSION
      9. REFERENCES
    3. 17. Finding Explicit and Implicit Knowledge: Biomedical Text Data Mining
      1. ABSTRACT
      2. INTRODUCTION
      3. BACKGROUND
      4. BIOMEDICAL TDM
        1. Automatic Gene Ontology Annotation
          1. Gene-Centered Document Representation
          2. Experiments
        2. Genetic Association Discovery
          1. Inference Network for Genetic Associations
          2. Probability Estimation
          3. Experimental Setup
          4. Results and Discussion
      5. FUTURE TRENDS
      6. CONCLUSION
      7. REFERENCES
      8. ENDNOTES
    4. 18. Rainstorm Forecasting By Mining Heterogeneous Remote Sensed Datasets
      1. ABSTRACT
      2. INTRODUCTION
      3. BACKGROUND
        1. Related Work
        2. Data Sources
          1. Satellite Images
          2. Satellite Observed Data
      4. MCS CLOUD DETECTION, TRACKING AND CHARACTERIZATION
      5. DATA PREPARATION
        1. Spatial Calibration
        2. Temporal adjustment
        3. Spatial Association
        4. Integrated Data Schema
      6. DECISION TREE BASED DATA MINING
      7. MCS CONCEPTUAL MODELING
      8. EXPERIMENTAL RESULTS AND ANALYSIS
      9. CONCLUSION
      10. ACKNOWLEDGMENT
      11. REFERENCES
    5. 19. Learning Verifiable Ensembles for Classification Problems with High Safety Requirements
      1. ABSTRACT
      2. INTRODUCTION
      3. SAFETY-RELATED SYSTEMS
        1. Assessing Solutions for Safety-Related Problems
        2. Overview of the Usage of Machine Learning in Safety-Related Domains
      4. LEARNING VERIFIABLE ENSEMBLES
        1. The Binary Classification Ensemble
          1. Projections of High-dimensional Data
          2. Submodel
          3. Learning the Verifiable Ensemble
          4. Applying the Verifiable Ensemble
          5. Discussion of the Learning Algorithm
        2. Safe Learning for Airbag Control
      5. EXTENSION TO MULTICLASS ENSEMBLES
        1. Common Multi-Class Extensions
          1. One-Against-Rest Multi-Class Extension
          2. One-Against-One Multi-Class Extension
          3. Risk of Inconsistent Decisions
        2. The Multi-Class Ensemble
          1. Hierarchy of Misclassification Costs.
          2. Ensemble of Multi-Class Submodels
          3. Hierarchical Separate-and-Conquer Ensemble
          4. One-Versus-Rest Ensemble
          5. Related Work
        3. A Medical Diagnosis Example
          1. Hierarchical Separate-and-Conquer Ensemble
          2. One-vs-Rest Ensemble
          3. Discussion of the Results
      6. DATA PREPROCESSING
        1. Data Filtering
          1. Convex Hull Filtering
          2. Upper Envelope Filtering
        2. Feature Construction
        3. A Naval Risk Detection Example
      7. CONCLUSION
      8. REFERENCES
  9. Compilation of References
  10. About the Contributors