Pattern Recognition in Practice IV: Multiple Paradigms, Comparative Studies and Hybrid Systems (eBook)

E.S. Gelsema, L.N. Kanal (Herausgeber)

eBook Download: PDF

2014 | 1. Auflage
586 Seiten
Elsevier Science (Verlag)
978-1-4832-9784-2 (ISBN)

Infos

The era of detailed comparisons of the merits of techniques of pattern recognition and artificial intelligence and of the integration of such techniques into flexible and powerful systems has begun.So confirm the editors of this fourth volume of Pattern Recognition in Practice, in their preface to the book.The 42 quality papers are sourced from a broad range of international specialists involved in developing pattern recognition methodologies and those using pattern recognition techniques in their professional work. The publication is divided into six sections: Pattern Recognition, Signal and Image Processing, Probabilistic Reasoning, Neural Networks, Comparative Studies, and Hybrid Systems, giving prospective users a feeling for the applicability of the various methods in their particular field of specialization.

Inhalt

Front Cover 1
Pattem Recognition in Practice IV: Multiple Paradigms, Comparative Studies and 
4 
Copyright Page 5
Table of Contents 12
PREFACE 6
ACKNOWLEDGEMENTS 10
PART I: 
18 
Chapter 1. 
20 
1. INTRODUCEN 20
2. PATTERN REPRESENTATION 21
3. PATTERN RELATIONS 22
4. 
23 
5. HUMAN INTERACTION 24
6. PROJECTS 25
7. CONCLUSION 26
APPENDIX 26
REFERENCES 27
Chapter 2. Application of evidence theory to 
30 
1. D-S THEORY 30
2. THE METHOD 32
3. SIMULATION RESULTS 36
4. CONCLUSION 41
REFERENCES 41
Chapter 3. 
42 
1. INTRODUCTION 42
2. OVERVIEW OF DECISION TREE METHODOLOGIES 43
3. A FRAMEWORK FOR DECISION TREE CONSTRUCTION 44
4. EXPERIMENTAL EVALUATION 48
5. CONCLUSIONS 51
REFERENCES 51
Chapter 4. 
54 
1. INTRODUCTION 54
2. STATISTICAL OBJECT RECOGNITION 55
3. HIDDEN MARKOV MODELS 55
4. OBJECT ORIENTED IMPLEMENTATION OF HMMS 57
5. AFFINE INVARIANT FEATURES 57
6. EXPERIMENTAL RESULTS 58
7. SUMMARY AND CONCLUSIONS 60
ACKNOWLEDGEMENT 60
REFERENCES 61
Chapter 5. 
62 
1. INTRODUCTION 62
2. CHUNKING 65
3. LOW ORDER MOMENT DESCRIPTORS 67
4. INFERENCE 68
5. GENERALISATION 70
6. INVARIANCE 70
7. NOISE 71
8. BINDING AND OCCLUSION 72
9. NEURAL MODELS 72
10. SUMMARY AND CONCLUSIONS 75
REFERENCES 75
Chapter 6. 
76 
1. INTRODUCTION 76
2. FINDING LINE SEGMENTS 78
3. FINDING SQUARES 80
4. FINDING CUBES 81
5. NOISE 82
6. INVARIANCE AND MANIFOLDS 83
7. OCCLUSION 87
8. CONCLUSION 90
REFERENCES 90
Chapter 7. 
92 
1. INTRODUCTION 92
2. METHOD 93
3. RESULTS 100
4. ROBUSTNESS 103
5. CONCLUSION AND FURTHER WORK 106
REFERENCES 106
Chapter 8. 
108 
1. INTRODUCTION 108
2. THE EM ALGORITHM 110
3. CONVERGENCE AND INITIAL CONDITIONS 111
4. CLUSTERING TECHNIQUES 112
5. THE DOG RABBIT STRATEGY 114
6. RESULTS 118
7. CONCLUSION AND SUMMARY 120
REFERENCES 122
Chapter 9. 
124 
1. INTRODUCTION 124
2. MODELING FRAMEWORK 127
3. METHODOLOGY 128
4. EXAMPLE 131
5. CONCLUSIONS 133
REFERENCES 133
Discussions Part I: 
136 
PART II: 
144 
Chapter 10. 
146 
1. INTRODUCTION 146
2. PIECE-WISE-LINEAR REGRESSION 147
3. MODEL SELECTION PROBLEM 149
4. APPLICATION TO IMAGE ANALYSIS 152
5. ROBUST REGRESSION AND HOUGH TRANSFORM 152
6. SUMMARY 156
REFERENCES 156
Chapter 11. 
158 
1. INTRODUCTION 158
2. REFLECTANCE RATIOS 160
3. RECOGNITION USING REFLECTANCE RATIOS 162
4. EXPERIMENTS 165
5. DISCUSSION 167
REFERENCES 169
Chapter 12. 
170 
1. INTRODUCTION 170
2. THE PROPOSED TECHNIQUE 171
3. EXPERIMENTAL RESULTS 178
4. CONCLUDING REMARKS 181
REFERENCES 181
Chapter 13. 
182 
1. INTRODUCTION 182
2. RELATONAL GRAPHS 184
3. MATCHING PROBABILITIES 185
4. EXPERIMENTS 188
5. CONCLUSIONS 191
REFERENCES 192
Chapter 14. 
194 
1. INTRODUCTION 194
2. INEXACT MATCHING USING HOPFIELD NETWORKS 196
3. EXPERIMENTATIONS 197
4. DISCUSSION AND CONCLUSIONS 199
REFERENCES 201
Chapter 15. 
202 
I. INTRODUCTION 202
2. INITIAL DATA AND PREPROCESSING 203
3. METHODS 204
4. PROPOSED SYSTEM 205
5. EXPERIMENTS AND RESULTS 209
6. CONCLUSION 210
REFERENCES 211
Chapter 16. 
214 
1. INTRODUCTION 214
2. PROBLEM SOLVING 215
3 MODEL INVERSION, HYPOTHESIS AND PARAMETER ESTIMATION 219
4 RADIOMETRIC MODELLING AND STATE DEFINITION 221
5 PARAMETER ESTIMATION 222
6 THE EXPERIMENT 224
7 CONCLUSIONS 226
8 REFERENCES 228
Chapter 17. 
230 
1. INTRODUCTION 230
2. SPECKLE NOISE ELIMINATION 231
3. TEXTURE FEATURES 234
4. IMAGE CLASSIFICATION AND SEGMENTATION 236
5. RESULT AND DISCUSSION 237
6. CONCLUSIONS AND FUTURE WORK 240
ACKNOWLEDGEMENT 241
REFERENCES 241
Discussions Part II: 
242 
PART II.: PROBABILISTIC 
248 
Chapter 18. 
250 
I. INTRODUCTION 250
2. FORMAL MODEL 250
3.TRACTABILITY 252
4. FUTURE WORK 256
REFERENCES 256
Chapter 19. 
258 
1. INTRODUCTION 258
2. PRELIMINARIES 258
3. RELATED APPROACHTES 259
4. EVALUATING DECISION NETWORKS 260
5. GENERATED STRATEGIES AND DECISION TREES 263
6. DECISION NETWORK POTENTIALS 268
7. CONCLUSION 269
REFERENCES 270
Chapter 20. 
272 
1. INTRODUCTION 272
2. GEON BASED RECOGNITION 273
3. BAYESIAN NETWORKS 274
4. BAYESIAN NETWORK FOR RECOGNITION 276
5. A CONTROL STRUCTURE FOR RECOGNITION 278
6. EXPERIMENTS 279
7. CONCLUSION 282
8. Acknowledgement 283
REFERENCES 283
Chapter 21. 
284 
1. INTRODUCTION 284
2. AN EXAMPLE 285
3. TARGET LANGUAGE 285
4. DISCRIMINATION RULES IN TARGET LANGUAGE 287
5. CHARACTERISTIC RULES IN TARGET LANGUAGE 290
ß. A GLOBAL PERSPECTIVE 293
7. SOME EXTENSIONS 294
8. CONCLUSION 294
ACKNOWLEDGMENT 295
References 295
Discussions Part III 296
PART IV: 
302 
Chapter 22. 
304 
1. INTRODUCTION 304
2. PARAMETERS ( WEIGHTS ) COMMON FOR ALL CLASSES 305
3. INFLUENCE OF A LOSS FUNCTION IN ANN TRAINING 308
4. INTRINSIC DIMENSIONALITY AND A GOOD SEPARABILITY OF THE CLASSES 311
5. DISCUSSION 313
ACKNOWLEDGMENT 314
REFERENCES 314
Chapter 23. 
316 
1. INTRODUCTION 316
2. BOLTZMANN MACHINES WITH RESTRICTED STATE SPACE 318
3. THE BOLTZMANN PERCEPTRON 320
4. JOINT PROBABILITY ESTIMATION 322
5. DISCUSSION 328
ACKNOWLEDGEMENTS 329
REFERENCES 329
Chapter 24. 
330 
1. INTRODUCTION 330
2. SYMBOLIC MAPPING OF A PERCEPTRON 331
3. SYMBOLIC ANALYSIS OF FEEDFORWARD NETWORKS 336
4. SUMMARY AND CONCLUSIONS 339
REFERENCES 340
Chapter 25. 
342 
1. INTRODUCTION 342
2. PROBLEM FORMULATION 344
3. PROBABILISTIC RELAXATION 347
4. NEURAL NET IMPLEMENTATION 349
5. CONCLUSIONS 351
REFERENCES 351
Chapter 26. 
354 
1. INTRODUCTION 354
2. THE SENSITISED PATH TRAINING SCHEME 355
3. RESULTS 358
4. DISCUSSION 361
REFERENCES 362
Chapter 27. 
364 
1. INTRODUCTION 364
2. ATTRIBUTE SELECTION 366
3. CONTRIBUTION ANALYSIS 366
4. APPLICATIONS 370
5. DISCUSSION 373
6. CONCLUSION 373
REFERENCES 374
Chapter 28. 
376 
1. INTRODUCTION 376
2. SOM IN COMPUTER VISION, CLUSTERING, AND VISUALIZATION 378
3. MLP IN COMPLEX SYSTEM MODELLING 380
4. SUMMARY 381
REFERENCES 381
Chapter 29. 
384 
1. INTRODUCTION 384
2. THE BASIC APPROACH 385
3. THE MAIN SET OF EXPERIMENTS 386
4. EXPERIMENTS ON EDGE-SHIFTING 392
5. CONCLUDING REMARKS 394
REFERENCES 395
Discussions Part IV 396
PART V: COMPARATIVE STUDIES 406
Chapter 30. 
408 
1. INTRODUCTION 408
2. THE NETTALK EXPERIMENTS 408
3. A COMPARISON OF THREE TRADmONAL DATA SETS 414
4. A COMPARISON OF THREE OTHER DATA SETS 416
5. CONCLUSIONS AND DISCUSSION 418
REFERENCES 419
Chapter 31. 
420 
1. INTRODUCTION 420
2. FEATURE SUBSET SEARCH ALGORITHMS 422
3. IMPLEMENTATION DETAILS 424
4. EXPERIMENTS 425
5. DISCUSSION 428
6. CONCLUDING REMARKS 429
REFERENCES 430
Chapter 32. 
432 
1. INTRODUCTION 432
2. BACKGROUND ON ALGORITHMS 433
3. EXPERIMENTAL SET-UP 434
4. RESULTS 434
5. DISCUSSION AND CONCLUSION 439
REFERENCES 440
Chapter 33. 
442 
1. INTRODUCTION 442
2. HEURISTIC METHODS FOR CLUSTERING 443
3. DESIGN CHOICES AND PARAMETER SELECTION 445
4. EXPERIMENTAL SETUP 446
5. RESULTS AND DISCUSSION 447
6. CONCLUSIONS 452
References 453
Chapter 34. 
454 
1. INTRODUCTION 454
2. CHARACTER DATABASE AND FEATURES 455
3. CLASSIFIERS 458
4. EXPERIMENTAL RESULTS 462
5. CONCLUSIONS 464
REFERENCES 464
Chapter 35. 
466 
1. INTRODUCTION 466
2. METHOD 467
3. EXPERIMENTAL RESULTS 471
4. CONCLUSIONS 476
REFERENCES 477
Discussions Part V 478
PART VI: 
488 
Chapter 36. 
490 
1. INTRODUCTION 490
2. TOWARDS CLASSIFIER-INDEPENDENT FEATURE SELECTION 491
3. RELATIVE FEATURE IMPORTANCE 492
4. GENETIC NEURAL FEATURE IMPORTANCE ESTIMATOR (GENFIE) 500
5. CONCLUSIONS 503
REFERENCES 503
Chapter 37. 
506 
1. INTRODUCTION 506
2. MULTISENSORY VISION SYSTEM 507
3. VISION PLANNER 510
4. IMPLEMENTATION AND EXPERIMENTAL RESULTS 514
5. DISCUSSION AND CONCLUSIONS 516
REFERENCES 517
Chapter 38. 
518 
1. INTRODUCTION 518
2. MOTIVATING EXAMPLE 519
3. THE HYBRID KNOWLEDGE BASE FRAMEWORK 520
4. MULTILEVEL INTEGRATION ARCHITECTURE 525
REFERENCES 529
Chapter 39. 
530 
1. INTRODUCTION 530
2. TRANSFORMATION PROCEDURE 531
3. HYBRID GENETIC ALGORITHM/NEURAL NETWORK PROCEDURE 536
4. EXPERIMENTAL RESULTS 538
5. CONCLUSIONS 540
REFERENCES 540
Chapter 40. 
542 
1. INTRODUCTION 542
2. DISCRIMINATION AND DETECTION OF FAULT CONDITIONS 542
3. HIDDEN MARKOV MODELS FOR ONLINE MONITORING 544
4. APPLICATION TO ANTENNA POINTING SYSTEM MONITORING 547
5. EXPERIMENTAL MODELS AND RESULTS 548
6. DISCUSSION OF RELATED WORK 551
7. CONCLUSION 552
REFERENCES 552
Chapter 41. 
554 
1. INTRODUCTION 554
2. THE MODEL 555
3. THE FRAMEWORK 556
4. NUSCRIPT 556
5. AI AND PATTERN RECOGNITION TECHNIQUES USED 558
6. THE KNOWLEDGE SOURCES 564
7. RESULTS 564
8. CONCLUSIONS AND FURTHER WORK 566
REFERENCES 566
Chapter 42. 
568 
1. INTRODUCTION 568
2. OUTLINE OF THE APPROACH 569
3. FORMALISMS FOR NEURAL AND SEMANTIC NETWORKS 570
4. APPLICATION 572
5. RESULTS 575
6. CONCLUSION 578
REFERENCES 579
Discussions Part VI 580
LIST OF AUTHORS 590
LIST OF KEYWORDS 592

Produkt-Details

• Erscheint lt. Verlag: 28.6.2014
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Informatik ► Netzwerke
• Themenwelt: Informatik ► Theorie / Studium ► Künstliche Intelligenz / Robotik
• ISBN-10: 1-4832-9784-5 / 1483297845
• ISBN-13: 978-1-4832-9784-2 / 9781483297842