1st Edition

Supervised and Unsupervised Pattern Recognition Feature Extraction and Computational Intelligence

Edited By Evangelia Miche Tzanakou Copyright 2000

    There are many books on neural networks, some of which cover computational intelligence, but none that incorporate both feature extraction and computational intelligence, as Supervised and Unsupervised Pattern Recognition does. This volume describes the application of a novel, unsupervised pattern recognition scheme to the classification of various types of waveforms and images.
    This substantial collection of recent research begins with an introduction to Neural Networks, classifiers, and feature extraction methods. It then addresses unsupervised and fuzzy neural networks and their applications to handwritten character recognition and recognition of normal and abnormal visual evoked potentials. The third section deals with advanced neural network architectures-including modular design-and their applications to medicine and three-dimensional NN architecture simulating brain functions. The final section discusses general applications and simulations, such as the establishment of a brain-computer link, speaker identification, and face recognition.
    In the quickly changing field of computational intelligence, every discovery is significant. Supervised and Unsupervised Pattern Recognition gives you access to many notable findings in one convenient volume.

    classifiers-an overview
    Criteria for optimal classifier design
    Categorizing the Classifiers
    Classifiers
    Neural Networks
    Comparison of Experimental Results
    System Performance Assessment
    Analysis of Prediction Rates from Bootstrapping Assessment
    ARTIFICIAL NEURAL NETWORKS: DEFINITIONS, METHODS, APPLICATIONS
    Definitions
    Training Algorithm
    Some Applications
    A SYSTEM FOR HANDWRITTEN DIGIT RECOGNITION
    Preprocessing of Handwritten Digit Images
    Zernike Moments (ZM) for Characterization of Image Patterns
    Dimensionality Reduction
    Analysis of Prediction Error Rates from Bootstrapping Assessment
    Summary
    OTHER TYPES OF FEATURE EXTRACTION METHODS
    Introduction
    Wavelets
    Invariant Moments
    Entropy
    Cepstrum Analysis
    Fractal Dimension
    Entropy
    SGLD Texture Features
    FUZZY NEURAL NETWORKS
    Pattern Recognition
    Optimization
    System Design
    Clustering
    APPLICATION TO HANDWRITTEN DIGITS
    Introduction to Character Recognition
    Data Collection
    Results
    Discussion
    Summary
    A UNSUPERVISED NEURAL NETWORK SYSTEM FOR VISUAL EVOKED POTENTIALS
    Data Collection and Preprocessing
    System Design
    Results
    Discussion
    CLASSIFICATION OF MAMMOGRAMS USING A MODULAR NEURAL NETWORK
    Methods and System Overview
    Modular Neural Networks
    Neural Network Training
    Classification Results
    The Process of Obtaining Results
    ALOPEX Parameters
    Generalization
    Conclusions
    "VISUAL OPHTHALMOLOGIST": AN AUTOMATED SYSTEM FOR CLASSIFICATION OF RETINAL DAMAGE
    System Overview
    Modular Neural Networks
    Applications to Ophthalmology
    Results
    Discussion
    A THREE-DIMENSIONAL NEURAL NETWORK ARCHITECTURE
    The Neural Network Architecture
    Simulations
    Discussion
    A FEATURE EXTRACTION ALGORITHM USING CONNECTIVITY STRENGTHS AND MOMENT INVARIANTS
    ALOPEX Algorithms
    Moment Invariants and ALOPEX
    Results and Discussion
    MULTILAYER PERCEPTRONS WITH ALOPEX: 2D-TEMPLATE MATCHING AND VLSI IMPLEMENTATION
    Multilayer Perceptron and Template Matching
    VLSI Implementation of ALOPEX
    IMPLEMENTING NEURAL NETWORKS IN SILICON
    The Living Neuron
    Neuromorphic Models
    Neurological Process Modeling
    SPEAKER IDENTIFICATION THROUGH WAVELET MULTIRESOLUTION DECOMPOSITION AND ALOPEX
    Multiresolution Analysis through Wavelet Decomposition
    Pattern Recognition with ALOPEX
    Methods
    Results
    Discussion
    FACE RECOGNITION IN ALZHEIMER'S DISEASE: A SIMULATION
    Methods
    Results
    Discussion
    SELF-LEARNING LAYERED NEURAL NETWORKS
    Neocognition and Pattern Classification
    Objectives
    Methods
    Study A
    Study B
    Summary and Discussion
    BIOLOGICAL AND MACHINE VISION
    Distributed Representation
    The Model
    A Modified ALOPEX Algorithm
    Application to Template Matching
    Brain-to-Computer Link
    Discussion
    Each section also has an introduction and references

    Biography

    Evangelia Miche Tzanakou

    "This book is an excellent source of knowledge of state-of-the-art feature extraction…Supervised and unsupervised learning and training schemes are notable finds…Exciting applications of signal and image analysis and recognition…This book provides in-depth guidance and inspiring ideas to new applications of signal and image analysis and recognition."
    --Tonglei Li, Ph.D., Purdue University, School of Pharmacy
    "…great efforts have been made in a number of communities to explore solutions to pattern recognition problems…this book describes their efforts made over ten researchers in the Neuroelectric and Neurocomputing Laboratories at Rutgers University. Along with concise introductory materials in pattern recognition, this volume presents several applications of supervised and unsupervised schemes to the classification of various types of signals and images…Unlike other books in neural networks, this book gives an emphasis on feature extraction as well, which provides a systematic way to deal with pattern recognition problems in terms of neural networks and computational intelligence…it is worth noting that each chapter contains an extensive bibliography that provides a reliable list of good references. We believe that readers will find this list very useful to understand the materials in the book and cautious beginners in the related fields might benefit from this list as well…helpful to a broad audience of graduate students, researchers, practicing engineers and professionals in computer and information science, electrical engineering, and biomedical informatics…this book reflects the long-term continuous endeavors of a research group for conducting innovatory researches, which could provide some useful hints to those novices in related fields…pioneering volume…welcomed by all interested in the fields of pattern recognition and computational intelligence…the editor's serious attempt to address the aforementioned issue must be welcomed by all interested in the fields of pattern recognition and computational intelligence and, therefore, this book deserves all credit."
    --Ke Chen, National Laboratory of Machine Perception and The Center for Information Science, Peking University, Beijing, China
    Promo Copy