Evolutionary Computation: Principles and Practice for Signal Processing / Edition 1

Evolutionary Computation: Principles and Practice for Signal Processing / Edition 1

by David B. Fogel
ISBN-10:
0819437255
ISBN-13:
9780819437259
Pub. Date:
07/28/2000
Publisher:
SPIE Press

Paperback - Rent for

Select a Purchase Option
  • purchase options

Temporarily Out of Stock Online


Overview

Evolutionary Computation: Principles and Practice for Signal Processing / Edition 1

Evolutionary computation is one of the fastest growing areas of computer science, partly because of its broad applicability to engineering problems. The methods can be applied to problems as diverse as supply-chain optimization, routing and planning, task assignment, pharmaceutical design, interactive gaming, and many others within the signal processing domain. The book is an outgrowth of successful SPIE short courses taught by the author. The examples span a range of applications and should be useful to a variety of readers with different backgrounds and expertise.

Product Details

ISBN-13: 9780819437259
Publisher: SPIE Press
Publication date: 07/28/2000
Series: Tutorial Texts in Optical Engineering Series
Pages: 184
Product dimensions: 7.34(w) x 9.66(h) x 0.45(d)

Table of Contents

Prefacexi
Chapter 1An Overview of Evolutionary Algorithms and Their Advantages1
1.1Introduction1
1.2Basics of Evolutionary Algorithms1
1.3Examples of Evolutionary Algorithm Applications3
1.4Advantages of Evolutionary Algorithms9
1.4.1Conceptual Simplicity9
1.4.2Broad Applicability11
1.4.3Outperform Classic Methods on Real Problems11
1.4.4Potential to Use Knowledge and Hybridize with Other Methods12
1.4.5Parallelism13
1.4.6Robust to Dynamic Changes13
1.4.7Capability for Self-Optimization14
1.4.8Able to Solve Problems with No Known Solutions15
1.5Principles and Practice of Evolutionary Algorithms16
References16
Chapter 2Evolving Models of Time Series19
2.1Introduction to Time Series Prediction and Modeling19
2.2Linear Models with Payoffs Other Than Least Squares22
2.2.1Examples of Evolving Polynomials for Arbitrary Criteria23
2.3Generating ARMA Models in One Step27
2.3.1The Traditional Two-Step Approach27
2.3.2Information Criteria for Model Building29
2.3.3Evolutionary Modeling31
2.4Neural Models36
2.4.1Evolving a Neural Predictor for a Chaotic Time Series40
2.4.2Evolving a Predictor of a Financial Time Series42
2.5Multiple Interactive Programs48
2.6Discussion50
References54
Chapter 3Evolutionary Clustering and Classification57
3.1Concepts of Clustering and Classification57
3.2Evolutionary Clustering57
3.2.1Method59
3.2.2Evaluating Solutions62
3.2.3Problems Investigated63
3.2.4Results and Discussion63
3.3Evolutionary Classification69
3.3.1Neural Networks for Classifying Sonar Data69
3.3.1.1Back Propagation70
3.3.1.2Simulated Annealing71
3.3.1.3Evolutionary Algorithms72
3.3.1.4Method and Materials72
3.3.1.5Experiments74
3.3.1.6Results78
3.3.2Neural Networks for Classifying Breast Cancer Data82
3.3.2.1Methods82
3.3.2.1.1Input Features and Neural Network Architecture83
3.3.2.1.2Training and Evaluation84
3.3.2.2Results85
3.3.2.3Discussion88
3.4Discussion90
References90
Chapter 4Evolving Control Systems95
4.1Introduction to Control95
4.2Evolutionary Control as Function Optimization95
4.2.1Three Textbook Problems95
4.2.1.1The Linear-Quadratic Problem96
4.2.1.2The Harvest Problem97
4.2.1.3The Push-Cart Problem98
4.2.1.4Experimental Results98
4.2.2Traffic Ramp Control98
4.2.2.1Background98
4.2.2.2Ramp-Metering Control Rules105
4.2.2.3Methods108
4.2.2.4Results109
4.2.2.4.1Scenario I: High Mainline and Ramp Demand109
4.2.2.4.2Scenario II: Moderate Mainline and Ramp Demand with an Incident112
4.2.2.4.3Scenario III: High Mainline and Ramp Demand with an Incident116
4.2.2.5Discussion118
4.3Evolutionary Control through Model Identification118
4.3.1Steps Toward Controlling Blood Pressure during Surgery118
4.3.1.1Introduction118
4.3.1.2Method, Materials, and Results119
4.3.1.3Discussion124
4.3.2Controlling a Pole Balanced on a Cart128
4.3.2.1System Dynamics128
4.3.2.2Method129
4.3.2.3Results131
4.3.2.4Discussion134
4.4Summary134
References136
Chapter 5Theory and Tools for Improving Evolutionary Algorithms139
5.1Theory139
5.1.1Binary Representations and Maximizing Implicit Parallelism139
5.1.2Crossover and Building Blocks140
5.1.3The Schema Theorem: The Fundamental Theorem of Genetic Algorithms143
5.1.4Proportional Selection and the K-Armed Bandit144
5.1.5A New Direction145
5.2Methods to Relate Parent and Offspring Fitness146
5.3Experiments with Fitness Distributions148
5.3.1Methods148
5.3.2Results152
5.4Discussion157
References159
Index163

Customer Reviews

Most Helpful Customer Reviews

See All Customer Reviews