Uncertainty in Artificial Intelligence (eBook)

Proceedings of the Eighth Conference (1992), July 17-19, 1992, Eighth Conference on Uncertainty in Artificial Intelligence, Stanford University

Bruce D'Ambrosio, Didier J. Dubois, Michael P. Wellman (Herausgeber)

eBook Download: PDF

2014 | 1. Auflage
378 Seiten
Elsevier Science (Verlag)
978-1-4832-8287-9 (ISBN)

Infos

Uncertainty in Artificial Intelligence: Proceedings of the Eighth Conference (1992) covers the papers presented at the Eighth Conference on Uncertainty in Artificial Intelligence, held at Stanford University on July 17-19, 1992. The book focuses on the processes, methodologies, technologies, and approaches involved in artificial intelligence. The selection first offers information on Relative Evidential Support (RES), modal logics for qualitative possibility and beliefs, and optimizing causal orderings for generating DAGs from data. Discussions focus on reversal, swap, and unclique operators, modal representation of possibility, and beliefs and conditionals. The text then examines structural controllability and observability in influence diagrams, lattice-based graded logic, and dynamic network models for forecasting. The manuscript takes a look at reformulating inference problems through selective conditioning, entropy and belief networks, parallelizing probabilistic inference, and a symbolic approach to reasoning with linguistic quantifiers. The text also ponders on sidestepping the triangulation problem in Bayesian net computations; exploring localization in Bayesian networks for large expert systems; and expressing relational and temporal knowledge in visual probabilistic networks. The selection is a valuable reference for researchers interested in artificial intelligence.

Inhalt

Front Cover 1
Uncertainty in Artificial 
4 
Copyright Page 5
Table of Contents 8
Preface 6
Acknowledgments 7
Chapter 1. ReS—A Relative Method for Evidential Reasoning 12
Abstract 12
1 Introduction 12
2 Our Point of View 13
3 An Example 17
4 Summary 19
Acknowledgements 19
References 19
Chapter 2. Optimizing Causal Orderings for Generating DAGs from Data 20
Abstract 20
1 Introduction 20
2 Preliminaries 20
3 The swap operator 22
4 The reversal operator 23
5 The cliquereunion operator 24
6 The unclique operator 25
7 The algorithm 25
8 Conclusion 27
Acknowledgement 27
References 27
Chapter 3. Modal Logics for Qualitative Possibility and Beliefs 28
Abstract 28
1 Introduction 28
2 A Modal Representation of Possibility 29
3 Beliefs and Conditionals 32
4 Concluding Remarks 34
Acknowledgements 35
References 35
Chapter 4. Structural Controllability and Observability in Influence Diagrams 36
Abstract 36
1 Introduction 36
2 Structural Observability in Influence Diagram 38
3 Structural Controllability in Influence Diagram 41
4 Conclusions 42
Acknowledgement 42
References 42
Chapter 5. Lattice-Based Graded Logic: A Multimodal Approach 44
Abstract 44
1· INTRODUCTION 44
2. THE LANGUAGE 45
3 SEMANTICS 48
4 RELATED WORK 49
CONCLUSION AND PERSPECTIVES 50
References 51
Chapter 6. Dynamic Network Models for Forecasting 52
Abstract 52
1 INTRODUCTION 52
2 DYNAMIC NETWORK MODELS 53
3 BUILDING AND REFINING A DNM 54
4 FORECASTING: INFERENCE WITH A DNM 55
5 SPECIAL DNMS 55
6 NUMERICAL EXAMPLE 56
7 RELATED WORK 58
8 CONCLUSION 58
Acknowledgments 58
References 58
Chapter 7. Reformulating Inference Problems Through Selective Conditioning 60
Abstract 60
1 INTRODUCTION 60
2 RELATED WORK 60
3 RANDOMIZED APPROXIMATION SCHEMES 61
4 RAS ALGORITHMS FOR INFERENCE 61
5 DEPENDENCE VALUE OF BELIEF NETWORKS 61
6 DEPENDENCE VALUE AND TRACTABILITY 62
7 PROBLEM REFORMULATION 62
8 DIRICHLET DISTRIBUTIONS 62
9 DIRICHLET STOPPING RULES 63
10 STRUCTURE AND EFFECTS OF PRIOR PROBABILITIES 63
11 ANALYSIS OF REFORMULATION TRADEOFFS 63
12 SUMMARY AND CONCLUSIONS 64
Acknowledgments 64
References 64
Chapter 8. Entropy and Belief Networks 66
Abstract 66
1 Introduction 66
2 Webs 66
3 Logarithmic Scores 67
4 An Alternative Model 68
References 69
Chapter 9. Parallelizing Probabilistic Inference Some Early Explorations 70
Abstract 70
1 Introduction 70
2 Background 70
3 Models 71
4 Method 73
5 Results 73
6 Discussion 75
7 Summary 77
References 77
Chapter 10. Objection-Based Causal Networks 78
Abstract 78
1 INTRODUCTION 78
2 OBJECTION-BASES STATES OF 
78 
3 COMPONENTS OF A CAUSAL NETWORK 80
4 FROM CAUSAL NETWORKS TO STATES OF BELIEF 83
5 DISCUSSION 83
Acknowledgement 83
References 84
Chapter 11. A Symbolic Approach to Reasoning with Linguistic Quantifiers 85
Abstract 85
1 INTRODUCTION 85
2 LATTICES OF LABELS 86
3 LOCAL PROPAGATION OF INTERVAL-VALUED PROBABILITIES 86
4 THE QUALITATIVE QUANTIFIED SYLLOGISM 87
5 ROBUSTNESS ANALYSIS 89
6 A QUALITATIVE ANALYSIS OF ADAMS' INFERENCE RULES 90
7 THE GENERALIZED BAYES THEOREM 91
8 SYMBOLIC CONSTRAINT PROPAGATION 92
9 CONCLUDING REMARKS 93
References 93
Chapter 12. Possibilistic Assumption based Truth Maintenance System, Validation in a Data Fusion Application 94
Abstract 94
1 A DATA FUSION APPLICATION 94
2 ATMS BACKGROUND 95
3 POSSIBILISTIC LOGIC 95
4 POSSIBILISTIC ATMS 96
5 COUPLING A II-ATMS AND AN 
97 
6 THE AGGREGATION PHASE IN SEFIR 97
7 RESULTS 99
8 A TEST SCENARIO 100
9 CONCLUSIONS 102
Acknowledgements 102
References 102
Chapter 13. An entropy-based learning algorithm of Bayesian conditional trees 103
Abstract 103
1 Introduction 103
2 Learning Conditional Trees 104
3 Learning Networks with Small Cutset of Root Nodes 105
4 Conditional trees and Similarity networks 105
5 Related optimization procedures 107
6 Summary 108
References 108
Chapter 14. Knowledge integration for conditional probability assessments 109
Abstract 109
1 INTRODUCTION 109
2 SOME PRELIMINARIES 110
3 COHERENCE OF (P, Q) 110
4 EXTENSION TO MARGINAL DISTRIBUTIONS 111
5 CONCLUSIONS 113
REFERENCES 114
Chapter 15. Integrating Model Construction and Evaluation 115
Abstract 115
1 Introduction 115
2 Review of ALTERID Language 116
3 Algorithm MCE 116
4 Example 119
5 Implementation 121
6 Future Directions 121
References 121
Chapter 16. Reasoning With Qualitative Probabilities Can Be Tractable 123
Abstract 123
1 Introduction: Infinitesimal 
123 
2 Preliminary Definitions: The Ranking k+ 125
3 Plausible Conclusions: Computing the Z+-rank 125
4 Belief Change, Soft Evidence, and Imprecise Observations 127
5 Conclusions 130
Acknowledgements 131
References 131
Chapter 17. A computational scheme for reasoning in dynamic probabilistic networks 132
Abstract 132
1 INTRODUCTION 132
2 TERMINOLOGY 133
3 REASONING IN DPNs 134
4 SUMMARY 139
Acknowledgements 140
References 140
Chapter 18. The Dynamic of Belief in the transferable belief model and Specialization-Generalization Matrices 141
Abstract 141
1. INTRODUCTION 141
2. THE TRANSFERABLE BELIEF MODEL 141
3. THE PRINCIPLE OF MINIMAL COMMITMENT 142
4. THE DYNAMIC OF THE TRANSFERABLE BELIEF MODEL 143
5. SPECIALIZATIONS 143
6. DEMPSTER'S RULES IN THE VIEW OF SPECIALIZATIONS 144
7. ANOTHER CHARACTERIZATION OF DEMPSTERS RULE OF COMBINATION 146
8. GENERALIZATION TO CONTRACTION OF BELIEFS AND DISJUNCTIVE COMBINATIONS 146
9. CONCLUSIONS 147
Acknowledgment 147
References 147
Chapter 19. A NOTE ON THE MEASURE OF DISCORD 149
Abstract 149
References 152
Chapter 20. Semantics for Probabilistic Inference 153
Abstract 153
1 INTRODUCTION 153
2. DEPARTURES FROM CONVENTIONAL ANALYSIS 153
3 IMPLICATION AND INDUCTIVE VALIDITY 154
4 SYNTAX 154
5 SEMANTICS 154
6 PROBABILISTIC SOUNDNESS 155
7 PREMISES 155
8 NUMBERS OF MODELS 155
9 SPECIFICITY 156
10 STRENGTH 156
11 CONFLICTING EVIDENCE 156
12 THE GENERAL CASE 156
13 INCONSISTENCY 157
14 RELATION TO OTHER WORK 157
15 CONCLUSION 158
Acknowledgement 158
References 158
Chapter 21. Some Problems for Convex Bayesians 160
Abstract 160
1 CONVEX BAYESIANISM 160
2 PROBLEMS 160
3 CONCLUSION 164
Acknowledgments 165
References 165
Chapter 22. Bayesian Meta-Reasoning: Determining Model Adequacy from Within a Small World 166
Abstract 166
1 INTRODUCTION 166
2 HYPOTHESIS TESTS AND THE BOUNDED BAYESIAN 166
3 A TEST STATISTIC 167
4 DETECTING INCORRECT STRUCTURE 168
5 INCOMPLETE DATA 169
6 DISCUSSION 169
Acknowledgments 169
References 169
Chapter 23. The Bounded Bayesian 170
Abstract 170
1 INTRODUCTION 170
2 THE SEQUENTIAL FORECASTING PROBLEM 171
3 MODEL SEARCH AND REVISION 171
4. INCORRECT MODELS 175
A APPENDIX: PROOFS OF RESULTS 175
Acknowledgments 176
References 176
Chapter 24. Representing Context-Sensitive Knowledge in a Network Formalism: A Preliminary Report 177
Abstract 177
1 INTRODUCTION 177
2 A PARTIAL NETWORK 178
3 REPRESENTATION OF CONCEPTS 178
4 STRUCTURE OF KNOWLEDGE BASE 181
5 INFERENCES SUPPORTED 181
6 SUPPORTING DECISION MAKING 182
7 DISCUSSION AND CONCLUSION 183
Acknowledgments 184
References 184
Chapter 25. A Probabilistic Network of Predicates 185
Abstract 185
1 INTRODUCTION 185
2 EVENT NETWORK 187
3 SCENARIOS AS EXPLANATIONS 188
4 INFERENCE ALGORITHM 189
5 PLAN RECOGNITION 189
6 IMPRECISE OBSERVATIONS 190
7 SPECIFICITY OF EXPLANATION 191
8 CONCLUSION 192
Acknowledgement 192
References 192
Chapter 26. Representing Heuristic Knowledge in D-S Theory 193
Abstract 193
1 INTRODUCTION 193
2 REPRESENTING HEURISTIC KNOWLEDGE IN D-S THEORY 194
3 THE RELATION BETWEEN EVIDENTIAL MAPPINGS AND BAYESIAN CONDITIONAL PROBABILITIES 197
4 CONSTRUCTING COMPLETE EVIDENTIAL MAPPING MATRICES TO PROPAGATE MASS FUNCTIONS FROM AN EVIDENCE SPACE TE TO A HYPOTHESIS SPACE 
198 
5 PROPAGATING BELIEFS USING HEURISTIC 
199 
6 CONCLUSION 200
Acknowledgement 201
References 201
Chapter 27. The Topological Fusion of Bayes Nets 202
Abstract 202
1 INTRODUCTION 202
2 COMPROMISE AND 
203 
3 TOPOLOGICAL FUSION 204
4 AN EXAMPLE 208
5 SUMMARY 209
References 209
Chapter 28. Calculating Uncertainty Intervals From Conditional Convex Sets of 
210 
Abstract 210
1 INTRODUCTION 210
2 PREVIOUS CONCEPTS AND 
211 
3 CONDITIONAL PROBABILITIES 
212 
4 TRANSFORMATION OF A CONDITIONAL PROBABILITY 
213 
5 CONCLUSIONS 217
Acknowledgements 217
References 217
Chapter 29. Sensor Validation using Dynamic Belief Networks 218
Abstract 218
1 INTRODUCTION 218
2 THE DOMAIN 219
3 INCORRECT DATA 220
4 HANDLING INCORRECT DATA WITHIN 
221 
5 EXPLAINING BAD DATA AS A 
223 
6 CONCLUSIONS 224
Acknowledgements 225
References 225
Chapter 30. Empirical Probabilities in Monadic Deductive Databases 226
Abstract 226
1 Introduction 226
2 Empirical Programs 226
3 Model Theoretic Semantics 227
4 Query Processing for Consistent 
230 
5 Related Work 232
6 Conclusions 233
Acknowledgements 233
References 233
Chapter 31. aHUGIN: A System Creating Adaptive Causal Probabilistic 
234 
Abstract 234
1 Introduction 234
2 Analysis of adaptation 234
3 Features of aHUGIN 236
4 Experiments with a HUGIN 237
References 240
Chapter 32. MESA: Maximum Entropy by Simulated Annealing 241
Abstract 241
1 INTRODUCTION 241
2 NOTATION AND COST 
242 
3 GENERATING A JOINT 
242 
4 MAXIMIZING THE ENTROPY 
243 
5 MARGINAL MODELS 244
6 ALGORITHM FOR MARGINAL 
244 
7 SUMMARY 246
Acknowledgements 246
References 246
8 APPENDIX 247
Chapter 33. Decision Methods for Adaptive Task-Sharing 
249 
Abstract 249
1 Introduction 249
2 Associate Systems 250
3 The Mars Rover Manager's Associate 
251 
4 Analysis and Review 253
5 Implications and Future Work 253
References 253
Chapter 34. Modeling Uncertain Temporal Evolutions 
255 
Abstract 255
1 INTRODUCTION 255
2 STOCHASTIC PROCESSES 
256 
3 DIAGNOSTIC FRAMEWORK 257
4 DISCUSSION 260
Acknowledgements 261
References 261
Chapter 35. Guess-And-Verify Heuristics for Reducing Uncertainties in 
263 
Abstract 263
INTRODUCTION 263
1. PROBLEM FORMULATION 263
2. EXACT AND HEURISTIC PROCEDURES 265
3. EXPERIMENTAL RESULTS AND A 
267 
4. CONCLUSIONS 268
References 268
Chapter 36. R& D Analyst: An Interactive Approach to Normative Decision System Model
270 
Abstract 270
1 INTRODUCTION 270
2 NORMATIVE DECISION SYSTEMS 270
3 BLACKBOARD ARCHITECTURES 271
4 AN OVERVIEW OF R& D ANALYST
5 AN OVERVIEW OF R& D ANALYST
6 ADVANCED ISSUES 276
7 RESEARCH ISSUES 277
8 CONCLUSIONS 277
Acknowledgements 277
References 277
Chapter 37. Possibilistic Constraint Satisfaction Problems or 
279 
Abstract 279
1 Introduction 279
2 Possibilistic constraint satisfaction 
280 
3 A design problem 284
4 Related works 285
5 Further researchs 285
Acknowledgements 285
References 285
Chapter 38. Decision Making Using Probabilistic Inference Methods 287
Abstract 287
1 INTRODUCTION 287
2 MAKING DECISIONS 287
3 USING GENERAL PROBABILISTIC 
288 
4 CLUSTERING ALGORITHM 
290 
5. DYNAMIC PROGRAMMING 292
6 CONCLUSIONS 293
7 ACKNOWLEDGEMENTS 294
8 REFERENCES 294
Chapter 39. Conditional Independence in Uncertainty Theories 295
Abstract 295
1 INTRODUCTION 295
2 VALUATION-BASED SYSTEMS 296
3 INDEPENDENCE AND 
299 
4 CONCLUSION 301
Acknowledgments 301
References 301
Chapter 40. The Nature of the unnormalized Beliefs 
303 
Abstract 303
1. INTRODUCTION 303
2. THE FRAME OF DISCERNMENT 304
3. CONDITIONING IN THE 
304 
4. UPDATING 
305 
5. THE EPISTEMIC CONSTRUCT OF 
307 
6. CONCLUSIONS 307
Acknowledgments 308
Bibliography 308
Chapter 41. Intuitions about Ordered Beliefs Leading to Probabilistic Models 309
Abstract 309
1 INTRODUCTION 309
2 ASSUMPTIONS FOR 
310 
3 FROM QUALITATIVE TO 
311 
4 A NOTE ON SET-VALUED 
311 
5 CONCLUSIONS 312
References 313
Chapter 42. Expressing Relational and Temporal Knowledge 
314 
Abstract 314
1 INTRODUCTION 314
2 EXPRESSING RELATIONAL 
314 
3 EXPRESSING TEMPORAL 
316 
4 APPLICATION TO COLONOSCOPY 317
5 CONCLUSIONS 320
Acknowledgments 320
References 320
Chapter 43. A Fuzzy Logic Approach to Target Tracking 321
Abstract 321
I. INTRODUCTION 321
II. PROBLEM FORMULATION 321
III. DESIGN OF THE FUZZY TRACKER 321
IV. SIMULATION RESULTS 323
V. CONCLUDING REMARKS 324
References 324
Chapter 44. Towards Precision of Probabilistic Bounds Propagation 326
Abstract 326
1 Introduction 326
2 The DUCK Calculus for Uncertain 
327 
3 Precise Bounds for Probabilistic 
329 
4 Summary and Outlook 332
References 333
Chapter 45. An Algorithm for Deciding if a Set of Observed Independencies 
334 
Abstract 334
1 Introduction 334
2 The DAG Construction Algorithm 335
3 Correctness 336
4 Complexity Analysis 337
5 Extensions and Improvements 337
Acknowledgement 338
Appendix: Proof of Lemmas 338
References 340
Chapter 46. Generalizing Jeffrey Conditionalization 342
Abstract 342
1 INTRODUCTION 342
2 GENERALIZED PROBABILITY 
342 
3 BOUNDING POSTERIORS BY 
343 
4 REVISING A PRIOR LOWER 
344 
5 OTHER APPROACHES 344
References 346
Chapter 47. INTERVAL STRUCTURE: 
347 
Abstract 347
1 INTRODUCTION 347
2 INTERVAL STRUCTURE INDUCED BY A 
348 
3 REPRESENTATIONS OF 
349 
4 KNOWLEDGE SYNTHESIS 
351 
5 CONCLUSION 353
References 354
Chapter 48. Exploring Localization In Bayesian Networks For 
355 
Abstract 355
1 LOCALIZATION 355
2 EXPLORE LOCALIZATION IN 
356 
3 MULTIPLY SECTIONED 
358 
4 TECHNICAL ISSUES 360
5 CONCLUSION 361
Acknowledgements 362
References 362
Chapter 49. A Decision Calculus for Belief Functions 
363 
Abstract 363
0. INTRODUCTION 363
1. VALUATION-BASED SYSTEM FOR 
364 
2. DECISION ANALYSIS USING 
365 
3. CONCLUSIONS 369
Acknowledgements 369
References 369
APPENDIX 369
Chapter 50. Sidestepping the Triangulation Problem 
371 
Abstract 371
1 INTRODUCTION 371
2 PREREQUISITE 372
3 DECOMPOSITION 373
4 PARALLEL REDUCTION 374
5 SERIAL REDUCTION 375
6 COMPONENT TREES 376
7 COMPONENT TREE PROPAGATION 377
8 CONCLUSIONS 378
Acknowledgement 378
References 378
Author Index 379

Produkt-Details

• Erscheint lt. Verlag: 12.5.2014
• Sprache: englisch
• Themenwelt: Informatik ► Theorie / Studium ► Künstliche Intelligenz / Robotik
• ISBN-10: 1-4832-8287-2 / 1483282872
• ISBN-13: 978-1-4832-8287-9 / 9781483282879