Engineering Evolutionary Intelligent Systems

Front Cover
Ajith Abraham, Crina Grosan, Witold Pedrycz
Springer Science & Business Media, Jan 3, 2008 - Computers - 444 pages
0 Reviews

Evolutionary design of intelligent systems is gaining much popularity due to its capabilities in handling several real world problems involving optimization, complexity, noisy and non-stationary environment, imprecision, uncertainty and vagueness. This edited volume 'Engineering Evolutionary Intelligent Systems' deals with the theoretical and methodological aspects, as well as various evolutionary algorithm applications to many real world problems originating from science, technology, business or commerce. This volume comprises of 15 chapters including an introductory chapter which covers the fundamental definitions and outlines some important research challenges. Chapters were selected on the basis of fundamental ideas/concepts rather than the thoroughness of techniques deployed.

 

What people are saying - Write a review

We haven't found any reviews in the usual places.

Contents

Methodologies Architectures and Reviews
1
2 Architectures of Evolutionary Intelligent Systems
2
3 Evolutionary Artificial Neural Networks
5
31 Evolutionary Search of Connection Weights
6
32 Evolutionary Search of Architectures
7
33 Evolutionary Search of Learning Rules
8
34 Recent Applications of Evolutionary Neural Networks in Practice
9
4 Evolutionary Fuzzy Systems
10
62 Modeling of Tungsten Inert Gas TIG Process
231
7 Concluding Remarks
248
Evolutionary Fuzzy Modelling for Drug Resistant HIV1 Treatment Optimization
251
12 Road Map
252
21 HIV Replication and Treatment Design
253
22 Experimental Settings and Data Collection
254
3 Machine Learning for Drug Resistant HIV
255
4 Fuzzy Modelling for HIV Drug Resistance Interpretation
256

41 Evolutionary Search of Fuzzy Membership Functions
12
43 Recent Applications of Evolutionary Fuzzy Systems in Practice
13
5 Evolutionary Clustering
14
6 Recent Applications of Evolutionary Design of Complex Paradigms
15
7 Multiobjective Evolutionary Design of Intelligent Paradigms
16
8 Conclusions
18
Analysis and Design of Rulebased Multilayer Perceptron Architectures
23
2 The architecture of conventional Hybrid Fuzzy Neural Networks HFNN
25
3 The architecture and development of genetically optimized HFNN gHFNN
27
32 Genetically optimized PNN gPNN
31
33 Optimization of gHFNN topologies
33
4 The algorithms and design procedure of genetically optimized HFNN gHFNN
34
in case of FS_FNN
35
in case of gPNN combined with FS_FNN
36
5 Experimental studies
40
52 Gas furnace process
45
53 NOx emission process of gas turbine power plant
49
6 Concluding remarks
53
7 Acknowledgement
55
Analysis and Design
58
1 Introduction
60
2 The architecture and development of the selforganizing neural networks SONN
62
22 Fuzzy Polynomial Neuron FPN based SONN and its topology
64
3 Genetic optimization of SONN
68
4 The algorithm and design procedure of genetically optimized SONN gSONN
70
5 Experimental studies
80
52 Chaotic time series
96
6 Concluding remarks
100
References
106
Evolution of Inductive Selforganizing Networks
109
2 Design of EAbased SOPNN
111
21 Representation of chromosome for appropriate information of each PD
112
22 Fitness function for modelling
115
3 Simulation Results
116
32 Threeinput nonlinear function
120
4 Conclusions
124
References
126
Recursive Pattern based Hybrid Supervised Training
129
11 Motivation
130
12 Organization of the chapter
132
22 Simplified architecture
133
24 Variable length genetic algorithm
134
25 Pseudo global optima
135
3 The RPHS training algorithm 31 Hybrid recursive training
137
32 Testing
139
5 The two spiral problem
140
6 Heuristics for making the RPHS algorithm better
141
61 Minimal coded genetic algorithms
142
63 Computation intensity and population size
143
64 Validation data
145
65 Early stopping
147
72 Experimental parameters and control algorithms implemented
149
73 Experimental results
150
8 Conclusions
154
References
155
Enhancing Recursive Supervised Learning Using Clustering and Combinatorial Optimization RSLCC
157
2 Some preliminaries
160
23 Related work
161
3 The RSLCC algorithm
163
32 Training
164
33 Simulation
165
41 Illustration
167
5 Heuristics for improving the performance of the RSLCC algorithm
169
52 Population size
170
56 Experimental parameters and control algorithms implemented
171
57 Results
172
6 Conclusions and future directions
175
Evolutionary Approaches to Rule Extraction from Neural Networks
177
2 The basics of neural networks
178
3 Rule extraction from neural networks
180
32 The existing methods of rule extraction from neural networks
182
4 Basic concepts of evolutionary algorithms
184
5 Evolutionary methods in rule extraction from neural networks
185
51 Local approach
186
52 Evolutionary algorithms in a global approach to rule extraction from neural networks
189
6 Conclusion
206
Clusterwise Design of Takagi and Sugeno Approach of Fuzzy Logic Controller
211
1 Introduction
213
2 Takagi and Sugeno Approach of FLC
215
3 Genetic Algorithm
216
4 Clustering and Linear Regression Analysis Using the Clustered Data
217
Clusterwise Linear Regression
219
5 GAbased Tuning of Takagi and Sugeno Approach of FLC
220
51 GeneticFuzzy System
221
6 Results and Discussion
223
41 Fuzzy Medical Diagnosis
258
42 Fuzzy Relational System for InVitro Cultures
259
43 Models for InVivo Clinical Data
260
5 Optimization Techniques
266
51 Fuzzy Genetic Algorithms
267
52 Random Searches
270
54 Feature Selection
271
6 Application
273
62 InVivo Prediction
276
63 Conclusions
283
7 Acknowledgements
284
References
285
A New Genetic Approach for Neural Network Design
288
2 Evolving ANNs
291
3 NeuroGenetic Approach
297
31 Evolutionary Algorithm
298
32 Individual Encoding
299
33 Fitness Function
300
34 Selection
301
35 Mutation
302
36 Recombination
304
4 RealWorld Applications
307
42 Brain Wave Analysis
312
43 Financial Modeling
315
5 Conclusion and Future Work
318
References
320
A Grammatical Genetic Programming Representation for Radial Basis Function Networks
325
2 Grammatical Evolution
326
3 Radial Basis Function Networks
328
4 GERBFN Hybrid
330
42 Example Individuals
331
6 Conclusions Future Work
333
References
334
Determining Waveinduced Seabed Liquefaction Depth
336
12 Genetic Algorithms
338
14 Waveinduced seabed liquefaction
339
2 A neuralgenetic technique for waveinduced liquefaction
341
21 Data preparation
342
3 Results and discussion
344
32 ANN model training using GAs for waveinduced liquefaction
346
33 Results for determining waveinduced liquefaction
347
4 Conclusions
350
On the Design of Largescale Cellular Mobile Networks Using Multipopulation Memetic Algorithms
353
2 Background and related work
356
3 Memetic Approach
359
32 Basic Principles of Memetic Algorithms
360
33 Multipopulation Approach
361
4 Implementation Details
362
42 Local Search Strategy
364
43 Experimental Setting
366
5 Performance Evaluation and Numerical Results
369
52 Quality of the Solutions
373
References
374
A Hybrid Cellular Genetic Algorithm for the Capacitated Vehicle Routing Problem
378
2 The Vehicle Routing Problem
382
3 The Proposed Algorithm
383
31 Problem Representation
385
32 Recombination
386
34 Local Search
387
the Way to JCell2o1i
389
41 Cellular vs Generational Genetic Algorithms
390
42 On the Importance of the Mutation Operator
391
43 Tuning the Local Search Step
393
5 Solving CVRP with JCell2o1i
395
51 Benchmark by Augerat et al
397
52 Benchmark by Van Breedam
398
53 Benchmark by Christofides and Eilon
400
54 Benchmark by Christofides Mingozzi and Toth
401
55 Benchmark by Fisher
403
57 Benchmark by Taillard
404
58 Benchmark of Translated Instances from TSP
406
7 Acknowledgement
407
A Best Found Solutions
410
B Results
411
Particle Swarm Optimization with Mutation for High Dimensional Problems
423
13 Curse of Dimensionality
424
2 PSO Modifications
425
22 Random Constriction Coefficient
426
31 Standard Test Functions
427
32 Neural Network Test Functions
429
4 Results and Discussion
431
41 Comparison with Differential Evolution
437
5 Conclusions and Future Work
438
Index
441
Copyright

Other editions - View all

Common terms and phrases

Bibliographic information