Matches in SemOpenAlex for { <https://semopenalex.org/work/W2478464795> ?p ?o ?g. }
Showing items 1 to 75 of
75
with 100 items per page.
- W2478464795 abstract "Explanation generation is an important and useful problem in automated reasoning. This thesis deals with the theoretical and practical issues involved in computing explanations in graphical models. Specifically, it deals with the problem of generating the Maximum a Posteriori Hypothesis (MAP), and a specialization of MAP known as the Most Probable Explanation (MPE). From a theoretical viewpoint, we analyze the complexity of producing explanations. For some classes it turns out to be very difficult. This hardness remains, even for classes of very simple models for which other inference tasks are tractable. Thus algorithms capable of producing good (though sometimes suboptimal) solutions efficiently are important. We give novel algorithms for approximating both MAP and MPE. While they provide no guarantee on the solution quality, in practice the performance seems to be quite good, and outperforms other previous algorithms. While approximations are useful, it is sometimes important to produce the optimal solution. We present a new MAP algorithm that is typically many orders of magnitude more efficient than previous techniques, and is the first algorithm capable of generating exact solutions for large real world models. We also introduce new theoretical and algorithmic results, which form the core of some of the approximation methods, but also have wider application. Specifically, we show how to extract additional information from jointree algorithms (the main class of algorithms for probabilistic inference), allowing them to directly answer more queries. We also introduce a novel jointree algorithm that improves the efficiency of computing such queries." @default.
- W2478464795 created "2016-08-23" @default.
- W2478464795 creator A5027252816 @default.
- W2478464795 creator A5089417455 @default.
- W2478464795 date "2003-01-01" @default.
- W2478464795 modified "2023-09-24" @default.
- W2478464795 title "Automated explanation generation in graphical models" @default.
- W2478464795 hasPublicationYear "2003" @default.
- W2478464795 type Work @default.
- W2478464795 sameAs 2478464795 @default.
- W2478464795 citedByCount "0" @default.
- W2478464795 crossrefType "journal-article" @default.
- W2478464795 hasAuthorship W2478464795A5027252816 @default.
- W2478464795 hasAuthorship W2478464795A5089417455 @default.
- W2478464795 hasConcept C105795698 @default.
- W2478464795 hasConcept C111472728 @default.
- W2478464795 hasConcept C11413529 @default.
- W2478464795 hasConcept C138885662 @default.
- W2478464795 hasConcept C154945302 @default.
- W2478464795 hasConcept C155846161 @default.
- W2478464795 hasConcept C2776214188 @default.
- W2478464795 hasConcept C2777212361 @default.
- W2478464795 hasConcept C2777472644 @default.
- W2478464795 hasConcept C2780586882 @default.
- W2478464795 hasConcept C33923547 @default.
- W2478464795 hasConcept C41008148 @default.
- W2478464795 hasConcept C49781872 @default.
- W2478464795 hasConcept C49937458 @default.
- W2478464795 hasConcept C75553542 @default.
- W2478464795 hasConcept C80444323 @default.
- W2478464795 hasConcept C9810830 @default.
- W2478464795 hasConceptScore W2478464795C105795698 @default.
- W2478464795 hasConceptScore W2478464795C111472728 @default.
- W2478464795 hasConceptScore W2478464795C11413529 @default.
- W2478464795 hasConceptScore W2478464795C138885662 @default.
- W2478464795 hasConceptScore W2478464795C154945302 @default.
- W2478464795 hasConceptScore W2478464795C155846161 @default.
- W2478464795 hasConceptScore W2478464795C2776214188 @default.
- W2478464795 hasConceptScore W2478464795C2777212361 @default.
- W2478464795 hasConceptScore W2478464795C2777472644 @default.
- W2478464795 hasConceptScore W2478464795C2780586882 @default.
- W2478464795 hasConceptScore W2478464795C33923547 @default.
- W2478464795 hasConceptScore W2478464795C41008148 @default.
- W2478464795 hasConceptScore W2478464795C49781872 @default.
- W2478464795 hasConceptScore W2478464795C49937458 @default.
- W2478464795 hasConceptScore W2478464795C75553542 @default.
- W2478464795 hasConceptScore W2478464795C80444323 @default.
- W2478464795 hasConceptScore W2478464795C9810830 @default.
- W2478464795 hasLocation W24784647951 @default.
- W2478464795 hasOpenAccess W2478464795 @default.
- W2478464795 hasPrimaryLocation W24784647951 @default.
- W2478464795 hasRelatedWork W1509251869 @default.
- W2478464795 hasRelatedWork W1587481346 @default.
- W2478464795 hasRelatedWork W1773349722 @default.
- W2478464795 hasRelatedWork W2079800412 @default.
- W2478464795 hasRelatedWork W2175676527 @default.
- W2478464795 hasRelatedWork W2176981155 @default.
- W2478464795 hasRelatedWork W2244358235 @default.
- W2478464795 hasRelatedWork W2247886904 @default.
- W2478464795 hasRelatedWork W2396464753 @default.
- W2478464795 hasRelatedWork W2473453452 @default.
- W2478464795 hasRelatedWork W2535143705 @default.
- W2478464795 hasRelatedWork W2729183986 @default.
- W2478464795 hasRelatedWork W2902934778 @default.
- W2478464795 hasRelatedWork W2903041470 @default.
- W2478464795 hasRelatedWork W2969357465 @default.
- W2478464795 hasRelatedWork W3027108890 @default.
- W2478464795 hasRelatedWork W3205094558 @default.
- W2478464795 hasRelatedWork W3214458206 @default.
- W2478464795 hasRelatedWork W149979836 @default.
- W2478464795 hasRelatedWork W2184069540 @default.
- W2478464795 isParatext "false" @default.
- W2478464795 isRetracted "false" @default.
- W2478464795 magId "2478464795" @default.
- W2478464795 workType "article" @default.