عنوان

Reasoning with probabilistic and deterministic graphical models :

(Number (ISBN

1681734907

(Number (ISBN

1681734915

(Number (ISBN

9781681734903

(Number (ISBN

9781681734910

Erroneous ISBN

1681734907

Erroneous ISBN

1681734923

Erroneous ISBN

9781681734903

Erroneous ISBN

9781681734927

Title Proper

Reasoning with probabilistic and deterministic graphical models :

General Material Designation

[Book]

Other Title Information

exact algorithms /

First Statement of Responsibility

Rina Dechter.

Edition Statement

Second edition.

Place of Publication, Distribution, etc.

[San Rafael, California] :

Name of Publisher, Distributor, etc.

Morgan & Claypool Publishers,

Date of Publication, Distribution, etc.

[2019]

Date of Publication, Distribution, etc.

©2019

Specific Material Designation and Extent of Item

1 online resource (xiv, 185 pages) :

Other Physical Details

illustrations

Series Title

Synthesis lectures on artificial intelligence and machine learning,

Volume Designation

#41

ISSN of Series

1939-4616 ;

Text of Note

Includes bibliographical references (pages 175-184).

Text of Note

1. Introduction -- 1.1 Probabilistic vs. deterministic models -- 1.2 Directed vs. undirected models -- 1.3 General graphical models -- 1.4 Inference and search-based schemes -- 1.5 Overview of the book

Text of Note

2. Defining graphical models -- 2.1 General graphical models -- 2.2 The graphs of graphical models -- 2.2.1 Basic definitions -- 2.2.2 Types of graphs -- 2.3 Constraint networks -- 2.4 Cost networks -- 2.5 Probability networks -- 2.5.1 Bayesian networks -- 2.5.2 Markov networks -- 2.6 Influence diagrams -- 2.7 Mixed networks -- 2.8 Summary and bibliographical notes

Text of Note

3. Inference: bucket elimination for deterministic networks -- 3.1 Bucket elimination for constraint networks -- 3.2 Bucket elimination for propositional CNFs -- 3.3 Bucket elimination for linear inequalities -- 3.4 The induced-graph and induced-width -- 3.4.1 Trees -- 3.4.2 Finding good orderings -- 3.5 Chordal graphs -- 3.6 Summary and bibliography notes

Text of Note

4. Inference: bucket elimination for probabilistic networks -- 4.1 Belief updating and probability of evidence -- 4.1.1 Deriving BE-bel -- 4.1.2 Complexity of BE-bel -- 4.1.3 The impact of observations -- 4.2 Bucket elimination for optimization tasks -- 4.2.1 A bucket elimination algorithm for mpe -- 4.2.2 A bucket elimination algorithm for map -- 4.3 Bucket elimination for Markov networks -- 4.4 Bucket elimination for influence diagrams -- 4.5 Bucket elimination for cost networks and dynamic programming -- 4.6 Bucket elimination for mixed networks -- 4.7 The general bucket elimination -- 4.8 Summary and bibliographical notes -- 4.9 Appendix: proofs

Text of Note

5. Tree-clustering schemes -- 5.1 Bucket-tree elimination -- 5.1.1 Asynchronous bucket-tree propagation -- 5.2 From bucket trees to cluster trees -- 5.2.1 From buckets to clusters -- the short route -- 5.2.2 Acyclic graphical models -- 5.2.3 Tree decomposition and cluster tree elimination -- 5.2.4 Generating tree decompositions -- 5.3 Properties of CTE for general models -- 5.3.1 Correctness of CTE -- 5.3.2 Complexity of CTE -- 5.4 Illustration of CTE for specific models -- 5.4.1 Belief updating and probability of evidence -- 5.4.2 Constraint networks -- 5.4.3 Optimization -- 5.5 Summary and bibliographical notes -- 5.6 Appendix: proofs

Text of Note

6. AND/OR search spaces for graphical models -- 6.1 AND/OR search trees -- 6.1.1 Weights of OR-AND arcs -- 6.1.2 Pseudo trees -- 6.1.3 Properties of AND/OR search trees -- 6.2 AND/OR search graphs -- 6.2.1 Generating compact AND/OR search spaces -- 6.2.2 Building context-minimal AND/OR search graphs -- 6.2.3 Size of AND/OR graph -- 6.3 Finding good pseudo-trees -- 6.3.1 Pseudo trees created from induced-graphs -- 6.3.2 Hypergraph decompositions -- 6.4 Value functions of reasoning problems -- 6.4.1 Searching and/or tree (AOT) and and/or graph (AOG) -- 6.5 General AND-OR search -- AO(i) -- 6.5.1 Complexity -- 6.6 AND/OR search algorithms for mixed networks -- 6.6.1 AND-OR-cpe algorithm -- 6.6.2 Constraint propagation in AND-OR-cpe -- 6.6.3 Good and nogood learning -- 6.7 Summary and bibliographical notes -- 6.8 Appendix: proofs

Text of Note

7. Combining search and inference: trading space for time -- 7.1 The cutset-conditioning scheme -- 7.1.1 Cutset-conditioning for constraints -- 7.1.2 General cutset-conditioning -- 7.1.3 Alternating conditioning and elimination -- 7.2 The super-cluster schemes -- 7.3 Trading time and space with AND/OR search -- 7.3.1 AND/OR cutset-conditioning -- 7.3.2 Algorithm adaptive caching (AOC.q/) -- 7.3.3 Relations between AOC(q), AO-ALT-VEC(q) and AO-VEC(q) -- 7.3.4 AOC(q) Compared with STCE(q) -- 7.4 Summary and bibliographical notes -- 7.5 Appendix: proofs

Text of Note

8. Conclusion -- Bibliography -- Author's biography.

0

8

8

8

8

8

8

8

Text of Note

Graphical models (e.g., Bayesian and constraint networks, influence diagrams, and Markov decision processes) have become a central paradigm for knowledge representation and reasoning in both artificial intelligence and computer science in general. These models are used to perform many reasoning tasks, such as scheduling, planning and learning, diagnosis and prediction, design, hardware and software verification, and bioinformatics. These problems can be stated as the formal tasks of constraint satisfaction and satisfiability, combinatorial optimization, and probabilistic inference. It is well known that the tasks are computationally hard, but research during the past three decades has yielded a variety of principles and techniques that significantly advanced the state of the art. This book provides comprehensive coverage of the primary exact algorithms for reasoning with such models. The main feature exploited by the algorithms is the model's graph. We present inference-based, message-passing schemes (e.g., variable-elimination) and search-based, conditioning schemes (e.g., cycle-cutset conditioning and AND/OR search). Each class possesses distinguished characteristics and in particular has different time vs. space behavior. We emphasize the dependence of both schemes on few graph parameters such as the treewidth, cycle-cutset, and (the pseudo-tree) height. The new edition includes the notion of influence diagrams, which focus on sequential decision making under uncertainty. We believe the principles outlined in the book would serve well in moving forward to approximation and anytime-based schemes. The target audience of this book is researchers and students in the artificial intelligence and machine learning area, and beyond.

Source for Acquisition/Subscription Address

01307586

Algorithms.

Bayesian statistical decision theory.

Graphical modeling (Statistics)

Machine learning.

Reasoning.

Algorithms.

Bayesian statistical decision theory.

COMPUTERS-- General.

Graphical modeling (Statistics)

Machine learning.

Reasoning.

COM-- 000000

Number

Edition

Class number

Book number

Dechter, Rina,1950-

Date of Transaction

20200823052156.0

Cataloguing Rules (Descriptive Conventions))

pn

Electronic name

[Book]

Y