2,696 Hits in 6.9 sec

Learning Bounded Treewidth Bayesian Networks

Gal Elidan, Stephen Gould
2008 Neural Information Processing Systems  
With the increased availability of data for complex domains, it is desirable to learn Bayesian network structures that are sufficiently expressive for generalization while also allowing for tractable inference  ...  In this work we present a novel method for learning Bayesian networks of bounded treewidth that employs global structure modifications and that is polynomial in the size of the graph and the treewidth  ...  Finally, we learn a bounded treewidth Bayesian network by iteratively augmenting the model with such chains.  ... 
dblp:conf/nips/ElidanG08 fatcat:tdll4l645vf5pdyeb4hneblhtm


2010 Biocomputing 2011  
and information learned from past subproblems which is facilitated through the use of depth-first branch and bound to explore the master search space.  ...  General pedigrees can be encoded as Bayesian networks, where the common MPE query corresponds to finding the most likely haplotype configuration.  ...  The structure of the AND/OR search tree is based on the underlying pseudo tree T : the root of the AND/OR search tree is an OR node labeled with the root of T .  ... 
doi:10.1142/9789814335058_0004 fatcat:cyekyq7e7ndudh25puuub3xsdy

Auto-Keras: An Efficient Neural Architecture Search System [article]

Haifeng Jin, Qingquan Song, Xia Hu
2019 arXiv   pre-print
In this paper, we propose a novel framework enabling Bayesian optimization to guide the network morphism for efficient neural architecture search.  ...  The framework develops a neural network kernel and a tree-structured acquisition function optimization algorithm to efficiently explores the search space.  ...  NETWORK MORPHISM GUIDED BY BAYESIAN OPTIMIZATION The key idea of the proposed method is to explore the search space via morphing the neural architectures guided by Bayesian optimization (BO) algorithm.  ... 
arXiv:1806.10282v3 fatcat:u4ieabmp7zgt7g2lpfvwv5gbzu

Learning Bounded Treewidth Bayesian Networks with Thousands of Variables [article]

Mauro Scanagatta, Giorgio Corani, Cassio P. de Campos, Marco Zaffalon
2016 arXiv   pre-print
We present a method for learning treewidth-bounded Bayesian networks from data sets containing thousands of variables.  ...  Bounding the treewidth of a Bayesian greatly reduces the complexity of inferences. Yet, being a global property of the graph, it considerably increases the difficulty of the learning process.  ...  As a second contribution, we propose two approaches for learning Bayesian networks with bounded treewidth.  ... 
arXiv:1605.03392v1 fatcat:66cp6fq3fzcn3kemwzxmcuhd7i

A Dynamic Approach for MPE and Weighted MAX-SAT

Tian Sang, Paul Beame, Henry A. Kautz
2007 International Joint Conference on Artificial Intelligence  
We consider the more general CNF-based MPE problem, where each literal in a CNF-formula is associated with a weight.  ...  The MPE-SAT algorithm is quite competitive with the state-of-the-art MAX-SAT, WCSP, and MPE solvers on a variety of problems.  ...  The induced width of a Bayesian network is a measure of its density.  ... 
dblp:conf/ijcai/SangBK07 fatcat:huqbcmrzwncqvnfmdhjjsjqq2e

Parameterized Complexity Results for Exact Bayesian Network Structure Learning

S. Ordyniak, S. Szeider
2013 The Journal of Artificial Intelligence Research  
Bayesian network structure learning is the notoriously difficult problem of discovering a Bayesian network that optimally represents a given set of training data.  ...  Results: We show that exact Bayesian network structure learning can be carried out in non-uniform polynomial time if the super-structure has bounded treewidth, and in linear time if in addition the super-structure  ...  In the following we will assume that we are given an instance I = (D, f, k) of k-O-Local Search Bayesian Network Structure Learning together with a nice tree decomposition (T, χ, r) for S f of width at  ... 
doi:10.1613/jair.3744 fatcat:4vxjfnzavnapxkujvl6eqqorhe

Best-First AND/OR Search for Most Probable Explanations [article]

Radu Marinescu, Rina Dechter
2012 arXiv   pre-print
The paper evaluates the power of best-first search over AND/OR search spaces for solving the Most Probable Explanation (MPE) task in Bayesian networks.  ...  The main contribution of this paper is in showing that a recent extension of AND/OR search algorithms from depth-first Branch-and-Bound to best-first is indeed very effective for computing the MPE in Bayesian  ...  Furthermore, we showed [3] that AOBB can improve its guiding heuristic function dynamically, by learning from portions of the search space that were already explored.  ... 
arXiv:1206.5268v1 fatcat:z4qd76htpfbpfiliatj7kuoge4

Toward parallel search for optimization in graphical models

Lars Otten, Rina Dechter
2010 International Symposium on Artificial Intelligence and Mathematics  
Subproblem generation is itself embedded into an AND/OR Branch and Bound algorithm and dynamically takes previous subproblem solutions into account.  ...  Based on the AND/OR graph search framework, the procedure exploits the structure of the underlying problem graph.  ...  its master search space through AND/OR graph search guided by the start pseudo tree T c .  ... 
dblp:conf/isaim/OttenD10 fatcat:irdvlyn2vffx5dbh2eifoyflsu

SAT-Based Approaches to Treewidth Computation: An Evaluation

Jeremias Berg, Matti Jarvisalo
2014 2014 IEEE 26th International Conference on Tools with Artificial Intelligence  
Malone, “Learning optimal bounded modifications to the encoding. Compared to earlier work, we treewidth bayesian networks via maximum satisfiability,” in Proc.  ...  treewidth Bayesian network structures [17].  ... 
doi:10.1109/ictai.2014.57 dblp:conf/ictai/BergJ14 fatcat:pcjlnsh4cvcqbnbb4zaelcadx4

Bayesian Inference in Monte-Carlo Tree Search [article]

Gerald Tesauro, V T Rajan, Richard Segal
2012 arXiv   pre-print
Monte-Carlo Tree Search (MCTS) methods are drawing great interest after yielding breakthrough results in computer Go.  ...  We further propose propagating inference in the tree via fast analytic Gaussian approximation methods: this can make the overhead of Bayesian inference manageable in domains such as Go, while preserving  ...  RELATED WORK Several extensions of UCT are known in the literature; most use the core UCT search strategy but provide alternative upper-bound functions to guide node selection.  ... 
arXiv:1203.3519v1 fatcat:egcrymgdxbe6hlhmm6c77hweni

Planning for Novelty: Width-Based Algorithms for Common Problems in Control, Planning and Reinforcement Learning [article]

Nir Lipovetzky
2021 arXiv   pre-print
Width-based algorithms search for solutions through a general definition of state novelty.  ...  Width-based algorithms performance is understood theoretically through the notion of planning width, providing polynomial guarantees on their runtime and memory consumption.  ...  Similar treewidth notions have been proposed to bound algorithms over intractable problems in constraint satisfaction and bayesian networks [Dechter, 2003; Pearl, 1988] . Definition 2.  ... 
arXiv:2106.04866v1 fatcat:yhsvlej5j5av3bf5iw7bwadvji

Learning Multivariate Distributions by Competitive Assembly of Marginals

Francisco Sanchez-Vega, Jason Eisner, Laurent Younes, Donald Geman
2013 IEEE Transactions on Pattern Analysis and Machine Intelligence  
The approach is motivated by compositional models and Bayesian networks, and designed to adapt to small sample sizes.  ...  We start with a large, overlapping set of elementary statistical building blocks, or "primitives", which are low-dimensional marginal distributions learned from data.  ...  gives a subset of Bayesian networks of tree-width w, while dropping the global constraint gives the superset CPCP w [46] .  ... 
doi:10.1109/tpami.2012.96 pmid:22529323 fatcat:annyoql4prhhnmlzc2n7utjz74

AND/OR Multi-Valued Decision Diagrams (AOMDDs) for Graphical Models

R. Mateescu, R. Dechter, R. Marinescu
2008 The Journal of Artificial Intelligence Research  
Inspired by the recently introduced framework of AND/OR search spaces for graphical models, we propose to augment Multi-Valued Decision Diagrams (MDD) with AND nodes, in order to capture function decomposition  ...  The first is search-based, and works by applying reduction rules to the trace of the memory intensive AND/OR search algorithm.  ...  This work was supported by the NSF grants IIS-0412854 and IIS-0713118, and the initial part by the Radcliffe fellowship 2005-2006 (through the partner program), with Harvard undergraduate student John  ... 
doi:10.1613/jair.2605 fatcat:pfoyf633kvaabkcsppwqm4b4sm

Stochastic Anytime Search for Bounding Marginal MAP

Radu Marinescu, Rina Dechter, Alexander Ihler
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
In this paper, we develop new search-based bounding schemes for Marginal MAP that produce anytime upper and lower bounds without performing exact likelihood computations.  ...  The empirical evaluation demonstrates the effectiveness of our new methods against the current best-performing search-based bounds.  ...  AND/OR Search Spaces AND/OR search is guided by a spanning pseudo tree of the graphical model's primal graph (in which any arc of the model not in the tree is a back-arc in the pseudo-tree) and the search  ... 
doi:10.24963/ijcai.2018/704 dblp:conf/ijcai/0002DI18 fatcat:xnx47pkymrg37fqbktpqx34yie

Approximate Inference by Compilation to Arithmetic Circuits

Daniel Lowd, Pedro M. Domingos
2010 Neural Information Processing Systems  
Arithmetic circuits (ACs) exploit context-specific independence and determinism to allow exact inference even in networks with high treewidth.  ...  We propose and evaluate a variety of techniques based on exact compilation, forward sampling, AC structure learning, Markov network parameter learning, variational inference, and Gibbs sampling.  ...  The WinMine Toolkit learns BNs with tree-structured CPDs, leading to complex models with high tree-width.  ... 
dblp:conf/nips/LowdD10 fatcat:3vugsjcgrnbndi6l7rftmnidwe
« Previous Showing results 1 — 15 out of 2,696 results