Swarm Intelligence in Data Mining

Size: px

Start display at page:

Download "Swarm Intelligence in Data Mining"

Meagan Patrick
5 years ago
Views:

1 Ajith Abraham Crina Grosan Vitorino Ramos (Eds.) Swarm Intelligence in Data Mining With 91 Figures and 73 Tables Springer

2 Contents 1 Swarm Intelligence in Data Mining Crina Grosan, Ajith Abraham and Monica Chis Biological Collective Behavior Swarms and Artificial Life Particle Swarm Optimization (PSO) Ant Colonies Optimization Data mining Steps of Knowledge Discovery Swarm Intelligence and Knowledge Discovery Ant Colony Optimization and Data mining Conclusions 16 References 16 2 Ante Constructing Ruie-Based Classifiers David Martens, Manu De Backer, RafHaesen, Bart Baesens, Tom Holvoet Introduction Ant Systems and Data Mining Ant Systems Data Mining Data Mining with Ant Systems AntMiner The Construction Graph Edge Probabilities Heuristic Value Pheromone Updating Early Stopping Distributed Data Mining With AntMiner+: a Credit Scoring Case Experiments and Results Experimental Set-Up Datasets 35

3 XIV Contents Credit Scoring 35 Toy Problems Software Implementation Discussion Conclusion and Future Research 40 References 41 3 Performing Feature Selection with ACO Richard Jensen Introduction Rough Feature Selection Theoretical Background Reduction Method Fuzzy-Rough Feature Selection Fuzzy Equivalence Classes Fuzzy Lower and Upper Approximations Fuzzy-Rough Reduction Method A Worked Example Ant-based Feature Selection ACO Framework Feature Selection 58 Selection Process 59 Complexity Analysis 59 Pheromone Update Crisp Ant-based Feature Selection Evaluation Experimental Setup Experimental Results Fuzzy Ant-based Feature Selection Evaluation..; Web Classification 63 System Overview 63 Experimentation and Results Systems Monitoring 66 Comparison of Fuzzy-Rough Methods 68 Comparison with Entropy-based Feature Selection 69 Comparison with the use of PCA 70 Comparison with the use of a Support Vector Classifier Conclusion 71 References 72 4 Simultaneous Ant Colony Optimization Algorithms for Learning Linguistic Fuzzy Rules Michelle Galea, Qiang Shen Introduction Background Fuzzy Rules and Rule-Based Systems 76

4 Contents XV Fuzzy Sets and Operators 77 Linguistic Variables and Fuzzy Rules 78 Classification using Fuzzy Rules 79 A Rule-Matching Example Ant Colony Optimization and Rule Induction Simultaneous Fuzzy Rule Learning Why Simultaneous Rule Learning FRANTIC-SRL 86 Rule Construction 86 Heuristic 87 Pheromone Updating 88 Transition Rule 88 Rule Evaluation Experiments and Analyses Experiment Setup 90 The Datasets 90 Other Induction Algorithms 91 FRANTIC-SRL Parameters Saturday Morning Problem Results Water Treatment Plant Results Conclusions and Future Work 95 References 97 5 Ant Colony Clustering and Feature Extraction for Anomaly Intrusion Detection Chi-Ho Tsang, Sam Kwong Introduction Related Works Ant Colony Clustering Model Basics and Problems of Ant-based Clustering Approach Measure of Local Regional Entropy Pheromone Infrastructure Modified Short-term Memory and oc-adaptation Selection Scheme, Parameter Settings and Cluster Retrieval Experiments and Results Dataset Description and Preprocessing Metrics of Cluster Validity and Classification Performance Cluster Analysis on Benchmark Datasets ACCM with Feature Extraction for Intrusion Detection Conclusions Future Works 121 References 121

5 XVI Contents 6 Particle Swarm Optimization for Pattern Recognition and Image Processing Mahamed G.H. Omran, Andries P. Engelbrecht, Ayed Salman Introduction Background The clustering problem 126 The AT-means Algorithm 128 The Fuzzy C-means Algorithm 129 Swarm Intelligence Approaches Color Image Quantization Spectral Unmixing 132 Linear Pixel Unmixing (or Linear Mixture Modeling) 132 Selection of the End-Members Particle Swarm Optimization A PSO-based Clustering Algorithm with Application to Unsupervised Image Classification Experimental Results A PSO-based Color Image Quantization (PSO-CIQ) Algorithm Experimental Results The PSO-based End-Member Selection (PSO-EMS) Algorithm The Generation of Abundance Images Experimental results Conclusion 148 References Data and Text Mining with Hierarchical Clustering Ants Hanene Azzag, Christiane Guinot, Gilles Venturini Introduction Biological and Computer modeis Ants based algorithms for clustering Self-assembly in real ants A Computer model of ants self-assembly for hierarchical clustering Self-assembly and robotics Two stochastic and deterministic algorithms Common principles Stochastic algorithm: AntTreesrocff Deterministic algorithm with no thresholds and no parameters : AntTree NO -THRESHOLDS Properties Experimental results with numeric, symbolic and textual databases Testing methodology Parameters study Tested algorithms Results with numeric databases Results with symbolic databases 168

6 Contents XVII Processing times Comparison with biomimetic methods Comparative study on textual databases Real world applications Human skin analysis Web usage mining Generation and interactive exploration of a portal site Incremental clustering of a large data set Principles of AntTree^c Results with incremental and large data sets Conclusions 186 References Swarm Clustering Based on Flowers PolUnation by Artincial Bees Majid Kazemian, Yoosef Ramezani, Caro Lucas, Behzad Moshiri Introduction Clustering What is clustering? Why swarm intelligence? Swarm clustering Some artificial modeis FPAB FPAB underlying algorithms 196 Picking up pollen 197 Pollinating 198 Natural selection 198 Merge algorithm Experimental results Conclusion and future works 200 References Computer study of the evolution of 'news foragers' on the Internet Zsolt Palotai, Sdndor Mandusitz, Andräs Lörincz Introduction Related work Forager architecture Algorithms Reinforcing agent Foragers Experimental results Environment Time lag and multiplication Compartmentalization Discussion Conclusions 217

7 XVin Contents References Data Swarm Clustering Christian Veenhuis, Mario Koppen Introduction Data Clustering Rock Algorithms Particle Swarm Optimization Data Swarm Clustering Initialization Iteration Cluster Retrieval Experimental Setup Synthetical Datasets Real Life Datasets Parameters Results Conclusion 240 References Clustering Ensemble Using ANT and ART Yan Yang, Mohamed Kamel, Fan Jin Introduction Ant Colony Clustering Algorithm with Validity Index (ACC-VI) Ant Colony Clustering Algorithm Clustering Validity Index ACC-VI Algorithm ART Algorithm Clustering Ensemble Model Consensus Functions ART Ensemble Aggregation Model Experimental Analysis Artificial Data Set (2D3C) Real Data Set (Iris) Reuter Document Collection Conclusions 262 Acknowledgements 262 References 262 Index 265

Particle Swarm Optimization Methods for Pattern. Recognition and Image Processing

Particle Swarm Optimization Methods for Pattern Recognition and Image Processing by Mahamed G. H. Omran Submitted in partial fulfillment of the requirements for the degree Philosophiae Doctor in the Faculty