The Practical Handbook of Genetic Algorithms: Applications, Second Edition

The Practical Handbook of GENETIC ALGORITHMS Applications......Page 1
Preface......Page 4
Contents......Page 13
Figures......Page 23
Tables......Page 36
0.1.1 Optimizing or Improving the Performance of Operating Systems......Page 40
0.1.4 Purpose of this Chapter......Page 41
0.2.2 Revising the Model or Revising the Data?......Page 42
0.2.4 Significance and Meaningfulness......Page 43
0.3.1 An Example:Linear Regression......Page 45
0.4 Iterative Hill-Climbing Techniques......Page 46
0.4.1 Iterative Incremental Stepping Method......Page 47
0.4.2 An Example:Fitting the Continents Together......Page 48
0.4.3 Other Hill-Climbing Methods......Page 50
0.4.4 The Danger of Entrapment on Local Optima and Saddle Points......Page 51
0.4.5 The Application of Genetic Algorithms to Model Fitting......Page 52
0.5.1 Description of the Problem......Page 53
0.5.2 A Model of Data Continuity......Page 54
0.5.3 Fitting the Data to the Model......Page 57
0.5.4 The Appropriate Misfit Function......Page 58
0.5.5.1 Model 0......Page 60
0.5.5.3 Model 2......Page 61
0.5.5.4 Comparison of Model 0, Model 1 and Model 2......Page 62
0.5.5.5 Interpretation of the Parameters......Page 63
0.5.6.1 The Genetic Algorithm Program......Page 64
0.5.6.4 Interpretation of the Results......Page 66
0.6 Conclusion......Page 67
Reference......Page 68
1.1 Introduction......Page 69
1.2 Fuzzy Modeling......Page 71
1.2.1 The Takagi-Sugeno Fuzzy Model......Page 72
1.2.2 Data-Driven Identification by Clustering......Page 73
1.3 Transparency and Accuracy of Fuzzy Models......Page 75
1.3.1 Rule Base Simplification......Page 76
1.3.2 Genetic Multi-objective Optimization......Page 77
1.4.1 Fuzzy Model Representation......Page 79
1.4.4 Crossover Operators......Page 80
1.4.5.1 Constraints......Page 81
1.5.1 Nonlinear Plant......Page 82
1.6 TS Singleton Model......Page 84
1.7 TS Linear Model......Page 87
1.7.1 Iris Classification Problem......Page 89
1.7.3 Proposed Approach......Page 90
References......Page 94
2.1 Introduction......Page 98
2.1.2 Previous Work......Page 99
2.2 Reorganization Operators......Page 100
2.2.1 The Jefferson Benchmark......Page 101
2.2.2 MTF......Page 104
2.2.3 SFS......Page 106
2.2.4 Competition......Page 110
2.3 The Experimentation......Page 111
2.3.1 The Languages......Page 112
2.3.2 Specific Considerations for the Language Recognition Problem......Page 115
2.4 Data Obtained from the Experimentation......Page 120
2.5 General Evaluation Criteria......Page 124
2.6.1 Machine Size......Page 125
2.6.2 Convergence Rates......Page 126
2.7 Conclusions and Further Directions......Page 128
References......Page 130
Appendix: Worksheets for L2, L2c and L8c.......Page 132
3.1 Introduction......Page 136
3.2.3 The Genetic Algorithm......Page 137
3.2.3.1 Genetic Algorithm Parameters......Page 142
3.3 Mapping the GA String into a Project Schedule and Computing the Fitness......Page 143
3.3.2 Imposing Constraints......Page 144
3.3.3 Calculation of Project Benefits......Page 146
3.3.3.2 Information Required......Page 147
3.3.3.3 Divisibility of User Benefits and Relationship to Travel Times......Page 148
3.3.4 Calculating Trip Generation, Route Choice and Link Loads......Page 151
3.4.1 Convergence of Solutions to the Problem......Page 154
3.4.2.1 The Ten Best Project Sequences......Page 156
3.4.2.2 Project Sequence Converted to Annual Investment......Page 157
3.4.2.3 Full Statement of the Best Project Schedule......Page 158
3.4.3 Similarity and Dissimilarity of Solutions: Euclidean Distance......Page 159
3.4.3.1 Similar Solutions with Close Objective Function Values......Page 165
3.4.3.3 Similar Solutions with Dissimilar Payoffs: The Shape of the Search Space......Page 166
3.5 Conclusions: Scheduling Interactive Road Projects by GA......Page 169
References......Page 170
4.1 Introduction......Page 172
4.2 Decoupled Regulator Configuration......Page 174
4.2.2 Chromosome and Population Structures......Page 176
4.3 Fitness Function for PCS......Page 177
4.3.1 OF 1 for Objective (1)......Page 178
4.3.2 OF 2 for Objective (2)......Page 180
4.4 Fitness function for FN......Page 181
4.4.2 OF 6 and OF 8 for Objective (2)and Objective (4)......Page 182
4.5 Steps of Optimization......Page 185
4.6 Design Example......Page 188
References......Page 202
5.1 Introduction......Page 204
5.2 Feature Selection......Page 206
5.3 Feature Selection by Genetic Algorithm......Page 207
5.3.1 GA Encoding Schemes......Page 208
5.3.3 GA Feature Selection Performance......Page 209
5.3.4 Conclusions......Page 210
5.4.1 Algorithm Design Issues......Page 211
5.4.2 Problem Representation......Page 212
5.4.4 Selection Strategy......Page 214
5.5.1 The Dataset......Page 215
5.5.2 Experiments on Two-Dimensional Data......Page 216
5.5.3 Results of Two-Dimensional Data Experiments......Page 217
5.5.5 Experiments on a Cell Image Dataset......Page 221
5.6.1 Parameterization Experiments......Page 223
5.6.2 Results of Parameterization Experiments......Page 224
5.6.3 Selecting the Neural Network Architecture......Page 225
5.7.1 Slide-Based vs.Cell-Based Features......Page 226
5.7.2 Comparison with the Standard Approach......Page 231
5.7.3 Discussion......Page 235
References......Page 236
6.1.1 Scope of This Chapter......Page 240
6.1.2 What is Multidimensional Scaling?......Page 241
6.1.2.2 Choice of the Misfit or Stress Function......Page 242
6.1.2.3 Choice of the Number of Dimensions......Page 243
6.1.2.5 Arbitrary Choice of Axes......Page 244
6.1.3.1 Limitations of the Standard Techniques......Page 245
6.2.1 A Simple One-Dimensional Example......Page 246
6.2.2 More than One Dimension......Page 248
6.2.3 Using Standard Multidimensional Scaling Methods......Page 250
6.3 A Genetic Algorithm for Multidimensional Scaling......Page 252
6.3.1.1 Binary and Real Parameter Representations......Page 253
6.3.1.2 Projected Mutation:A Hybrid Operator......Page 254
6.3.2.1 Inter-object Crossover......Page 255
6.3.4 Design and Use of a Genetic Algorithm for Multidimensional Scaling......Page 256
6.4.1 Systematic Projection......Page 258
6.4.2 Using the Genetic Algorithm......Page 259
6.4.3 A Hybrid Approach......Page 260
6.5.1 The Extend Model......Page 262
6.5.2.1 Within the Control Panel (Dialog Box)......Page 263
6.5.3 The Main Program......Page 264
6.5.4.1 CHECKVALIDATA()......Page 265
6.5.4.4 DESCEND()......Page 266
6.5.4.8 CROSSOBJ()......Page 267
6.5.5.3 Changes to the Main Program......Page 268
6.6 Using the Extend Program......Page 269
References......Page 270
Abstract......Page 271
7.1.2 GAB Approach for Single-Objective Bilevel Programming Models......Page 272
7.1.2.1 Bilevel Programming Problems......Page 273
7.1.2.2 GAB Approach......Page 274
7.1.2.3 Numerical Example......Page 276
7.1.3 GAB Approach for Multi-Objective Bilevel Programming Models......Page 280
7.1.3.1 Multi-Objective Bilevel Models......Page 281
7.1.3.3 Numerical Example......Page 282
7.1.4 Summary......Page 286
7.2.2 Review of TFS......Page 287
7.2.3 Calibration Measures......Page 289
7.2.4 GAB Calibration Procedure......Page 292
7.2.5 Calibration of TFS......Page 293
7.2.6 Case Study......Page 294
7.2.6.1 Path Choice Entropy and NCV:the Best Measures for Calibration......Page 296
7.2.6.2 Calibration Results......Page 299
7.3 Concluding Remarks......Page 303
References......Page 304
Appendix I: Notation......Page 307
8.1 Introduction......Page 310
8.2 The Job-Shop Scheduling Constraint Satisfaction Problem......Page 311
8.3 The Genetic Algorithm......Page 312
8.4 Fitness Refinement......Page 314
8.4.1 Variable and Value Ordering Heuristics......Page 315
8.5 Heuristic Initial Population......Page 317
8.6 Experimental Results......Page 319
8.7 Conclusions......Page 326
References......Page 327
9.1 Introduction......Page 329
9.2.1 Modeling the Conceptual Design Process......Page 330
9.3 The Implicit Redundant Representation Genetic Algorithm......Page 331
9.4 The IRR Genotype/Phenotype Representation......Page 333
9.4.1 Provision of Dynamic Redundancy......Page 335
9.4.2 Controlling the Level of Redundancy in the IRR GA Initial Population......Page 336
9.5.1 Unstructured Design Problem Formulation......Page 337
9.5.2.1 Encoding Frame Members as Design Variables......Page 338
9.5.2.2 Definition of Member Data Structures......Page 339
9.5.2.3 Creation of Linked Lists of Pointers Using SaveNodes()......Page 340
9.5.2.4 Construction of Complete Design Synthesis Alternatives......Page 344
9.5.3 Use of Repair Strategies on Frame Design Alternatives......Page 345
9.5.3.2 Deletion of Single Nodes on Floor Levels......Page 346
9.5.3.3 Replacement of Geometrically Similar Nodes with a Single Node......Page 348
9.5.4 Generation of Horizontal Members in Design Synthesis Alternatives......Page 351
9.5.5 Specification of Loads on Unstructured Frame Design Alternatives......Page 353
9.5.5.1 Specification of Gravity Loads......Page 354
9.5.5.2 Specification of Wind Loads......Page 355
9.6.1 Statement of Frame Design Objectives Used as Fitness Functions......Page 357
9.6.2.1 Stress Penalty Function......Page 359
9.6.2.2 Deflection Penalty Functions......Page 360
9.6.2.3 Symmetry Penalty Functions......Page 362
9.6.2.4 Composite Fitness Function......Page 364
9.7.1 Fitness Sharing among Individuals in the Population......Page 365
9.7.2 Tournament Selection of New Population Individuals......Page 366
9.7.3 Multiple Point Crossover of Binary Strings......Page 367
9.8 Results of the Implicit Redundant Representation Frame Synthesis Trials......Page 368
9.8.2 Synthesis versus Optimization of Frame Design Solutions Using IRR GA......Page 369
References......Page 373
Abstract......Page 375
10.2 Constraint Handling in EAs......Page 376
10.3.1 Heuristic Genetic Operators......Page 379
10.3.2 Knowledge-Based Fitness and Genetic Operators......Page 380
10.3.3 Glass-Box Approach......Page 381
10.3.4 Genetic Local Search......Page 382
10.3.5 Co-evolutionary Approach......Page 383
10.3.7 Stepwise Adaptation of Weights......Page 384
10.4 Discussion......Page 385
10.5 Assessment of EAs for CSPs......Page 386
10.6 Conclusion......Page 389
References......Page 390
11.1 Introduction......Page 396
11.2 FLC for the Boost Rectifier......Page 398
11.2.1.Switching Rule for the Switch SW......Page 399
11.2.2.1 Fuzzification......Page 400
11.2.3 Defuzzification......Page 403
11.3.2.1 The peak point......Page 404
11.3.3 Formulation of Multi-objective Fitness Function......Page 408
11.3.5 Crossover and Mutation Operations......Page 409
11.3.6 Validation of SI :Recovery of Valid Fuzzy Subsets......Page 411
11.4 Illustrative Example......Page 412
References......Page 422
12.1 Introduction......Page 424
12.2 Intelligent Control......Page 425
12.3.1 Optimal Control Concept......Page 426
12.3.2 Process Stability during Genetic Algorithm Optimizing......Page 429
12.3.3 Optimizing Criteria......Page 430
Solution representation.......Page 432
Algorithm termination conditions.......Page 433
12.5 Laboratory Cascaded Plant......Page 434
Heat exchanger 2.......Page 444
12.6.1 Non-coordinated Multilevel Control Using a PID Controller......Page 445
12.7 Fuzzy Multilevel Coordinated Control......Page 452
12.7.1 Decision Control Table......Page 454
12.8 Conclusions......Page 467
References......Page 469
13.1.Introduction and Chapter Overview......Page 475
13.2.1 Radiation Therapy Treatment Planning (RTP)......Page 476
13.2.2 Volumes......Page 477
13.2.3 Treatment Planning......Page 478
13.2.4 Recent Developments and Areas of Active Research......Page 479
13.2.4.1 Treatment Modalities......Page 480
13.2.4.2 Prediction of Treatment Outcome......Page 481
13.2.5 Treatment Planning......Page 483
13.2.4.4 Optimisation of Beam Orientation......Page 485
13.2.4.5 Optimisation Using Simulated Annealing......Page 486
13.3 Evolutionary Artificial Neural Networks......Page 487
13.3.1 Evolving Network Weights......Page 488
13.3.2 Evolving Network Architectures......Page 490
13.3.3 Evolving Learning Rules......Page 492
13.3.4 EPNet......Page 493
13.3.5 Addition of Virtual Samples......Page 494
13.4.1 The Backpropogation ANN for Treatment Planning......Page 495
13.4.2 Development of an EANN......Page 498
13.4.3 EANN Results......Page 503
13.4.4 Breast Cancer Treatment Planning......Page 509
13.5 Summary......Page 512
13.6 Discussion and Future Work......Page 514
Acknowledgments......Page 516
References......Page 517
Abstract......Page 521
14.2 A Heuristic Method......Page 522
14.3.1 Encoding......Page 524
14.3.3 Fitness Evaluation......Page 525
14.4.1 Heuristic Method......Page 527
14.4.2 Genetic Algorithm Based Solutions......Page 528
References......Page 534