Symbolic and Quantitative Approaches to Reasoning with Uncertainty

Name: Symbolic and Quantitative Approaches to Reasoning with Uncertainty | 11th European Conference, ECSQARU 2011, Belfast, UK, June 29-July 1, 2011, Proceedings
Brand: Springer
Price: 96.29 EUR
Availability: OnlineOnly

11th European Conference, ECSQARU 2011, Belfast, UK, June 29-July 1, 2011, Proceedings

Weiru Liu(Editor)

Springer (Publisher)

Published on 25. June 2011

XIV, 762 pages

E-Book

PDF with digital watermarking

System requirements

978-3-642-22152-1 (ISBN)

€96.29incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Intro
Title Page
Preface
Organization
Table of Contents
Invited Talks
Information Fusion and Revision in Qualitative and Quantitative Settings Steps Towards a Unified Framework
Introduction
Epistemic States
Plausible Inference vs. Prioritized Combination
Revision vs. Fusion
Elementary Revision
Elementary Fusion
Revision and Fusion across Uncertainty Theories
Revision and Fusion of Comparative Possibilities
Revision and Fusion of Probability Measures
Revision and Fusion of Possibility Distributions
Belief Functions
Conclusion
References
Introducing Equational Semantics for Argumentation Networks
Introduction
Aims of This Paper
Equational Examples
Formal Theory of the Equational Approach to Argumentation Networks
Numerical Calculations
Equational Approach to Logic
Conclusion
References
Constructive Decision Theory: Short Summary
References
Argumentation
Strong Equivalence for Argumentation Semantics Based on Conflict-Free Sets
Introduction
Background
Characterizations for Strong Equivalence
Strong Equivalence wrt. Naive Semantics
Strong Equivalence wrt. Stage Semantics
Strong Equivalence wrt. cf2 Semantics
Relation between Different Semantics wrt. Strong Equivalence
Comparing Semantics wrt. Strong Equivalence
Strong Equivalence and Symmetric Frameworks
Local Equivalence
Conclusion
References
Backing and Undercutting in Defeasible Logic Programming
Introduction
Background and Motivation
The Extended Representational Language
Arguments and Defeaters
Conclusions and Related Work
References
Arguing with Valued Preference Relations
Introduction
Argumentation Theory
Dung's Argumentation Framework
Preference-Based Argumentation Framework
Argumentation Framework with Varied-Strength Defeats
Valued Preference-Based Argumentation Framework
Link between the Defense for Boolean and Valued Preference Relations
General Inclusion Result between the Defense for Boolean and Valued Preference Relations
Case When P Derives from a Valuation on the Arguments
From a Valuation of Arguments to a Valued Preference Relation
Study of $P^w_1$
General Properties of $P_w$
Related Work
Conclusion
References
Arguing about the Trustworthiness of the Information Sources
Introduction
Meta-Argumentation
Modelling Trust in Meta-Argumentation
Information Sources
Evidence for Arguments
Focused Trust Relationships
Related Work and Conclusions
References
Two Roles of Preferences in Argumentation Frameworks
Introduction
Basics of Argumentation
Preferences in Argumentation: Informal Discussion
Handling Critical Attacks
Refining AFs by Preferences
A Particular Rich PAF
Related Work
Conclusion
References
Bayesian Networks and Causal Networks
A Re-definition of Mixtures of Polynomials for Inference in Hybrid Bayesian Networks
Introduction
Mixture of Polynomials Functions
MOP Functions
Fitting MOPs to Two- and Three-Dimensional CLG PDFs
Measuring Goodness of Fit of Approximations
One-Dimensional CLG Distributions
Two-Dimensional CLG Distributions
Three-Dimensional CLG Distributions
Summary and Discussion
References
Using Four Cost Measures to Determine Arc Reversal Orderings
Introduction
Definitions
Bayesian Networks
Arc Reversal
Four Cost Measures for AR Child Orderings
AR Child Orderings Using Four Cost Measures
Analysis
Experimental Results
Conclusions
References
Using the Noisy-OR Model Can Be Harmful . . . But It Often Is Not
Introduction
Preliminaries
The Noisy-OR Model
Empirical Studies Involving the Noisy-OR Model
The Propagation Effects of the Noisy-OR Model
Propagation in Causal Direction
Propagating in Diagnostic Direction
Revisiting Our Results in View of Noisy-OR Generalizations
Conclusions and Future Work
References
Attaining Monotonicity for Bayesian Networks
Introduction
Bayesian Networks
Monotonicity in Bayesian Networks
Reducing the Graphical Structure
The Intersection-of-Intervals Method
The Method
Applying the Intersection-of-Intervals Method
Complexity of the Intersection-of-Intervals Method
Some Practical Considerations
Conclusions
References
Importance Sampling on Bayesian Networks with Deterministic Causalities
Introduction
Importance Sampling with CVT
Importance Sampling and Sample Rejection
The CVT Approach
CVT Construction and Sample Matching
Results
MSE Results
Sample Rejection Reduction Results
CVT Compared to Related Work
Measuring the CVT Overhead
Conclusions
References
Bayesian Networks and the Imprecise Dirichlet Model Applied to Recognition Problems
Introduction
BNs, Dynamic BNs and Parameter Learning
Constraints
Imprecise Dirichlet Model
Maximum Entropy
The Learning Algorithm
Experiments
Facial Action Unit Recognition
Activity Recognition
Conclusion
References
On Stopping Evidence Gathering for Diagnostic Bayesian Networks
Introduction
Preliminaries
Bayesian Networks
Complexity Theory
The Stop Problem
Complexity Results
Practical Considerations
Conclusions
References
SemCaDo: A Serendipitous Strategy for Learning Causal Bayesian Networks Using Ontologies
Introduction
Basic Concepts and Background
Causal Bayesian Networks
Ontologies
Related Work
SemCaDo: A Serendipitous Causal Discovery Algorithm for Ontology Evolution
CBNs vs. Ontologies
SemCaDo Sketch
Toy Example
Experimental Study
Conclusions and Future Works
References
Scaling Up the Greedy Equivalence Search Algorithm by Constraining the Search Space of Equivalence Classes
Introduction
Learning Bayesian Networks by Using Greedy Equivalence Search
The E-Space
Greedy Equivalent Search (GES)
Scaling Up GES
GESG
GESiC
GESiCG
Experimental Evaluation
Empirical Results
Conclusions
References
Extensions of Decision-Theoretic Troubleshooting: Cost Clusters and Precedence Constraints
Introduction
Troubleshooting Scenarios
Basic Troubleshooting
Troubleshooting with Postponed System Test
Troubleshooting with Cost Clusters without Inside Information
Complexity Results
Reduction
Troubleshooting with Cost Clusters without Inside Information
Troubleshooting with Precedence Constraints
Conclusions and Future Research
References
Locally Averaged Bayesian Dirichlet Metrics
Introduction
Bayesian Score Metrics
Averaged Bayesian Dirichlet Metrics
Motivation
Locally and Globally Averaged Bayesian Dirichlet Metrics
Experiments
Experimental Settings
Conclusions and Future Works
References
Mixture of Markov Trees for Bayesian Network Structure Learning with Small Datasets in High Dimensional Space
Introduction
Bayesian Network Structure Learning in High Dimension
Introduction
MMHC Algorithm
Mixture of Bayesian Networks Structured Trees
MtMHC Algorithm
MtMHC Principle
MtMPC Algorithm
MtMHC Optimization
Empirical Simulations and Results
Experimental Protocol
Results
Summary and Future Works
References
Finding P-Maps and I-Maps to Represent Conditional Independencies
Introduction
Graphoids and Fast Closure
Graph Representation
The BN-Draw Procedure
The DAG-Representability Theorem
The Algorithm PC*
Finding an I-Map
An Algorithm to Find $G(J*)$
Conclusions
References
Marginalization without Summation Exploiting Determinism in Factor Algebra
Introduction
Formal Preliminaries
Variable Elimination as Factor Rewriting
Factor Indexing
Empirical Evaluation
Conclusions and Future Work
References
Belief Functions
Independence and 2-Monotonicity: Nice to Have, Hard to Keep
Introduction
Preliminaries
Lower Expectations and Credal Sets
2-Monotone Lower Probability and Möbius Inverse
2-Monotonicity Preservation Under Independence Assumptions
Strong Independence
Epistemic Irrelevance
Epistemic Independence
A 2-Monotone Outer-Approximation
Definition and Basic Properties
Discussion about Practical Interest
Conclusions
References
Constructing Dynamic Frames of Discernment in Cases of Large Number of Classes
Introduction
Handwriting Word Recognition
Background
DST Combination of HMM Classifiers
Computational Issues
Dynamic Frames of Discernment
Strategy 1: Intersection
Strategy 2: Union
Strategy 3: Borda Count
Strategy 4: Probabilistic Pre-processing
Experiments and Results
Datasets and HMM Classifiers
Results and Discussion
Conclusion
References
On Consistent Approximations of Belief Functions in the Mass Space
Introduction
Semantics of Consistent Belief Functions
The $L_p$ Consistent Approximation Problem
Consistent Approximation in M
L1 Approximation
L8 Approximation
L2 Approximation
Comparison with Approximation in the Belief Space
Conclusions
References
Generalized Information Theory Based on the Theory of Hints
Introduction
The Theory of Hints
The Pignistic Entropy
Pignistic Entropy and Aggregate Uncertainty
The Hints Entropy
Hints Entropy and Aggregate Uncertainty
Conclusion
References
Towards an Alarm for Opposition Conflict in a Conjunctive Combination of Belief Functions
Introduction
Belief Function Theory: Basic Concepts
Representing Information
Discounting Information
Pignistic Transformation
Distance between Two Belief Functions
Combining Different Pieces of Information
Conjunctive Rules of Combination
Disjunctive Rule of Combination
Combination with Adapted Conflict (CWAC)
With Two Belief Functions
General Case
Properties
Results
Example 1
Example 2
Conclusion and Future Work
References
E2GK: Evidential Evolving Gustafsson-Kessel Algorithm for Data Streams Partitioning Using Belief Functions
Introduction
Background
Gustafson-Kessel Algorithm
Belief Functions and Credal Partition
ECM: Evidential C-Means Algorithm
Deriving E2GK
E2GK: Evidential Evolving Gustafsson-Kessel Algorithm
Application of E2GK
Conclusion
References
Evidential Markov Decision Processes
Introduction
Existing Extensions of the MDP Model
The Classical MDP Model
The EMC Model
Existing Generalizations of the MDP
The Proposed Approach
The New EMDP Model
Link with the MDPST Model
Solving Method for EMDPs
Illustrative Example
Context of the Application Example
Obtained Results
Conclusion
References
Continuous Belief Functions to Qualify Sensors Performances
Introduction
Belief Function Framework
Belief on Real Numbers
Maximum of Necessity
Model of Sensor
Belief Functions Induced by Sampling
Experimental Settings
The Dempster's Approach
Likelihood and Maximum of Necessity Principles
Combination of Information
Independant Sources of Information
Inference Using the Max
Results
Conclusion
References
Belief Revision and Inconsistency Handling
Measuring Consistency Gain and Information Loss in Stepwise Inconsistency Resolution
Introduction
Preliminary Definitions
Inconsistency and Information Measures
Inconsistency Measures for Knowledgebases
Information Measures for Knowledgebases
Stepwise Inconsistency Resolution
Inconsistency Resolution by Deletion
Inconsistency Resolution by Weakening
Inconsistency Resolution Using Splitting
Discussion
References
Relating Truth, Knowledge and Belief in Epistemic States
Introduction
Notation and Terminology
KBR-Structures
Motivation
Definition of $K B_R$-Structures
Epistemic Properties of KB_R-Structures
S4.2 Representation of $KB_R$-Structures
Related Work - Further Research
References
How Strong Can an Agent Believe Reported Information ?
Motivation
A Model of Reported Information in Modal Logic
First Case: One Level of Nesting
Second Case: Two Levels of Nesting
General Case
A Model of Reported Information in the Theory of Evidence
First Case: One Agent
Second Case: Two Agents
General Case
Discussion
References
Logic-Based Fusion of Complex Epistemic States
Preliminaries
Epistemic States Fusion Operators
Rationality Postulates
Faithful Assignments
Representation Theorems
Fusion in a Concrete Representation of Epistemic States
A Concrete Example
Conclusion
References
Classification and Clustering
Latent Tree Classifier
Introduction
Latent Tree Classifier
A Learning Algorithm
Model Selection
Model Search
Empirical Evaluation
Experimental Settings
Classification Accuracy
Discovery of Latent Structures
Main Findings
Detailed Analysis
Related Work
Conclusions
References
When Learning Naive Bayesian Classifiers Preserves Monotonicity
Introduction
Preliminaries
Bayesian Networks and Classifiers
Qualitative Concepts of Probability
Monotonicity of Bayesian Networks
An Example of Counterintuitive Behaviour
Incorrect Monotonicity Directions
Investigating the Example
Deriving Correct Monotonicity Directions
Classification Behaviour
Related Work
Conclusions and Further Research
References
Possibilistic Classifiers for Uncertain Numerical Data
Introduction
Related Works
Possibilistic Classification
Possibilistic Distributions for Imperfect Numerical Data
Processing of Uncertain Classes in the Training Set
Processing of Imprecise Attributes in the Testing Set
Individual Possibility Distribution
Probability to Possibility Transformation-Based Classifiers
Proximity Based Classifiers
Experiments and Discussion
Generation of Imperfect Data
Classification Results
Conclusion
References
Default Reasoning and Logics for Reasoning under Uncertainty
Relational Probabilistic Conditional Reasoning at Maximum Entropy
Introduction
Relational Probabilistic Knowledge Representation
Default Reasoning Properties - System $P^prob$
Semantics for Relational Probabilistic Conditional Logic
Grounding Semantics with Constraints
Averaging Semantics
Aggregating Semantics
Relational Probabilistic Conditional Reasoning
Related Work
Conclusions and Discussion
References
Probabilistic Approach to Nonmonotonic Consequence Relations
Introduction
Preliminaries
Hyperreal Numbers
Nonstandard Probability Measures
Preferential Relations
Hierarchy of Preferential Relations
Probabilistic Representations of Inference Relations
Conditional Excluding Middle
Determinacy Preservation
Fragmented Disjunction
The e, µ Preferential Relation
Concluding Remarks
References
Bridging the Gap between Reinforcement Learning and Knowledge Representation: A Logical Off- and On-Policy Framework
Introduction
Preliminaries
Normal Logic Programs
Q-Learning and SARSA
Action Language $B_Q$
Language Syntax
Semantics
Off- and On-Policy Model-Free Reinforcement Learning Using Answer Set Programming
Correctness
Conclusions and Related Work
References
Answer Set Programming for Computing Decisions Under Uncertainty
Introduction
Qualitative Decision in Stratified Propositional Bases
Making Decision in ASP
LPODs and Abduction
Fully Certain Knowledge and All-or-Nothing Preferences
Computation of Optimal Pessimistic Decisions
Making Decision Under Uncertainty in ASP
Basic Definitions of LPPODs
Uncertain Knowledge and Prioritized Preferences
Classical ASP-Based Computation
Possibilistic ASP-Based Computation
Related Work
Concluding Remarks
References
Quasi Conjunction and Inclusion Relation in Probabilistic Default Reasoning
Introduction
Some Preliminary Notions
Probabilistic Entailment and Quasi Conjunction
Further Results on Quasi Conjunction and Inclusion Relation
Conclusions
References
Handling Exceptions in Logic Programming without Negation as Failure
Introduction
Background and Notation
Handling Default Reasoning
Ordering Rules by Specificity
Rules Rewriting
Consistency Test
Minimal Model Computation
Related Work
Concluding Remarks
References
Probabilistic Stit Logic
Introduction
The Base Logic: XSTITp
Choice with a Bounded Chance of Success
Conclusion and Discussion
References
Overriding Subsuming Rules
Introduction
Rules with Exceptions
Defaults
PEC Rules
Reasoning with and about PEC Rules
A Versatile Approach
X-Derivations
PEC-Implicants
Overriding Subsuming Rules
Conclusions and Future Work
References
Foundations of Reasoning and Decision Making under Uncertainty
Pseudo-polynomial Functions over Finite Distributive Lattices
Introduction and Motivation
Preliminaries
Characterization and Factorization of Pseudo-polynomial Functions
An Example
Pseudo-polynomial Functions over Chains
References
A Bridge between Probability and Possibility in a Comparative Framework
Introduction
Preliminaries
Coherent Extensions
Weakly Logically Independent Partitions
Possibility Measures as Enlargement of a Coherent Probability
Comparative Degree of Belief
Extending Comparative Probabilities
Conclusion
References
Leximax Relations in Decision Making through the Dominance Plausible Rule
Preliminaries
Three Characterizations of Leximax
Characterization for the Relations Defined by DPR Using a Leximax
Conclusion
References
Parameterized Uncertain Reasoning Approach Based on a Lattice-Valued Logic
Introduction
Lattice-Valued First-Order Logic
Uncertain Reasoning Approach Based on Lattice- Valued Logic Lvfl
Parameter Selection
An Illustrative Example
Conclusions
References
Fuzzy Sets and Fuzzy Logic
From Preference Relations to Fuzzy Choice Functions
Introduction
Preliminaries
Crisp Relations
Fuzzy Relations
From Complete Preference Relations to Fuzzy Choice Functions
Conclusions
References
Fuzzy Relational Inequations and Equations in the Framework of Control Problems
Introduction
Preliminaries
Closedness of Fuzzy Sets Under a Fuzzy Relation
Closedness and Relational Inequations
Conclusion
References
Fuzzy Autoepistemic Logic: Reflecting about Knowledge of Truth Degrees
Introduction
Background
Autoepistemic Logic
Fuzzy Logics
Fuzzy Autoepistemic Logic
Relation between Fuzzy Answer Set Programming and Fuzzy Autoepistemic Logic
Related Work
Conclusions
References
Belief Functions on MV-Algebras of Fuzzy Events Based on Fuzzy Evidence
Introduction
Preliminary Notions
Belief Functions on Boolean Algebras
MV-Algebras and States
Belief Functions on Finite Domain MV-Clans
Belief Functions and Their Integral Representations
Combining Belief Functions
Conclusion and Future Work
References
Order Compatible Fuzzy Relations and Their Elicitation from General Fuzzy Partitions
Introduction
Basic Definitions
General Fuzzy Partitions
Orderly Gradual Fuzzy Relations
Generating OCFRs from GFPs
Conclusion
References
Implementation and Applications of Uncertain Systems
Web Services and $Incerta Spiriti$ :A Game Theoretic Approach to Uncertainty
Introduction
Orc and Uncertainty Profiles
$Angels and Daemons as Web Incerta Spiriti$
Partial Order Sets
Uncertainty and Risk
Conclusions
References
Underwater Archaeological 3D Surveys Validation within the Removed Sets Framework
Introduction
The Validation Problem in VENUS
The Removed Sets Framework
Removed Sets Fusion
Partially Preordered Removed Sets Inconsistency Handling
ASP Implementation
The Validation Problem within RSF
The Validation Problem within PPRSIH
Concluding Discussion
References
Adaptive Dialogue Strategy Selection through Imprecise Probabilistic Query Answering
Introduction
Preliminaries
Dialogue Systems
Using Probabilistic Logic Programs to Represent Imprecise Data
Observation vs. a Priori Facts
The Experiment and Evaluation of Our Framework
Conducting the Experiment and Constructing a PLP
Evaluating the Results
Related Work and Conclusion
References
Possibility Theory and Possibilistic Logic
Statistical Estimations of Lattice-Valued Possibilistic Distributions
Introduction
Statistical Estimations of Lattice-Valued Possibility Degrees - a Formalized Model
Asymptotic Properties of Statistical Estimations of Upper-Semilattice-Valued Possibilistic Degrees
Some More Results on Upper Semilattice-Valued Lattices
Conclusions
References
Compiling Min-based Possibilistic Causal Networks: A Mutilated-Based Approach
Introduction
Basic Backgrounds on Possibility and Compilation
Possibility Theory
Possibilistic Causal Networks
Compilation Concepts
Causal Inference Using ?-DNNFs
Did We First Mutilate the Network?
Inference Process
Causal Inference Using Compiled Possibilistic Bases
Conclusion
References
Possibilistic Evidence
Introduction
A Maxitive Weighted Subset Representation Framework
Set Functions Associated with Possibilistic Evidence
Duality Relations
Particular Cases of Basic Possibility Assignment
Possibilistic Evidence as Imprecise Possibilities
Non Unicity of the Basic Possibility Assignment
Entailment between Basic Possibility Assignments
Definitions
Behavior of Cores and Supports with Respect to Entailment
Combination Laws
The Definitions
Properties of the Combination Rule
Concluding Remarks
References
Uncertainty in Databases
A Preference Query Model Based on a Fusion of Local Orders
Introduction
Related Approaches
Pareto-Order-Based Approaches
Outranking-Based Approach
A New Approach Based on a Fusion of Local Orders
Basic Version
Taking Scoring Functions into Account
Gradual Extension
Query Processing
Conclusion
References
Approximate Achievability in Event Databases
Introduction
Preliminaries on Event DBs
Optimal State Change Attempts
Different Kinds of Effect Estimators
Trie-enhanced Optimal State Change Attempt (TOSCA)
Experimental Evaluation
Related Work and Conclusions
References
A Probabilistic Interpretation for a Geometric Similarity Measure
Introduction
Related Work
CQQL Retrieval Model
Probabilistic Interpretation
Conclusion and Outlook
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Symbolic and Quantitative Approaches to Reasoning with Uncertainty

Description

More details

Other editions

Additional editions

Content

System requirements