DOCSLIB.ORG

On the Parameterized Complexity of Polytree Learning

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
On the Parameterized Complexity of Polytree Learning On the Parameterized Complexity of Polytree Learning Niels Grüttemeier , Christian Komusiewicz and Nils Morawietz∗ Philipps-Universität Marburg, Marburg, Germany {niegru, komusiewicz, morawietz}@informatik.uni-marburg.de Abstract In light of the hardness of handling general Bayesian networks, the learning and inference prob- A Bayesian network is a directed acyclic graph that lems for Bayesian networks fulfilling some spe- represents statistical dependencies between vari- cific structural constraints have been studied exten- ables of a joint probability distribution. A funda- sively [Pearl, 1989; Korhonen and Parviainen, 2013; mental task in data science is to learn a Bayesian Korhonen and Parviainen, 2015; network from observed data. POLYTREE LEARN- Grüttemeier and Komusiewicz, 2020a; ING is the problem of learning an optimal Bayesian Ramaswamy and Szeider, 2021]. network that fulfills the additional property that its underlying undirected graph is a forest. In One of the earliest special cases that has received atten- this work, we revisit the complexity of POLYTREE tion are tree networks, also called branchings. A tree is a LEARNING. We show that POLYTREE LEARN- Bayesian network where every vertex has at most one par- ING can be solved in 3n · |I|O(1) time where n is ent. In other words, every connected component is a di- the number of variables and |I| is the total in- rected in-tree. Learning and inference can be performed in stance size. Moreover, we consider the influence polynomial time on trees [Chow and Liu, 1968; Pearl, 1989]. of the number of variables d that might receive a Trees are, however, very limited in their modeling power nonempty parent set in the final DAG on the com- since every variable may depend only on at most one other plexity of POLYTREE LEARNING. We show that variable. To overcome this problem, a generalization of branchings called polytrees has been proposed. A poly- POLYTREE LEARNING has no f(d) · |I|O(1)-time algorithm, unlike Bayesian network learning which tree is a DAG whose underlying undirected graph is a for- est. An advantage of polytrees is that the inference task can be solved in d O(1) time. We show that, in 2 · |I| can be performed in polynomial time on them [Pearl, 1989; contrast, if d and the maximum parent set size are Guo and Hsu, 2002]. POLYTREE LEARNING, the problem bounded, then we can obtain efficient algorithms. of learning an optimal polytree structure from parent scores, however, remains NP-hard [Dasgupta, 1999]. We study exact 1 Introduction algorithms for POLYTREE LEARNING. Bayesian networks are the most important tool for modelling statistical dependencies in joint probability distributions. A Related Work. POLYTREE LEARNING is NP-hard even if every parent set with strictly positive score has size arXiv:2105.09675v1 [cs.DS] 20 May 2021 Bayesian network consists of a DAG (N, A) over the vari- [ ] able set N and a set of condiditional probability tables. Given at most 2 Dasgupta, 1999 . Motivated by the contrast a Bayesian network and the observed values on some of its between the NP-hardness of POLYTREE LEARNING and variables, one may infer the probability distributions of the the fact that learning a tree has a polynomial-time algo- remaining variables under the observations. One of the draw- rithm, the problem of optimally learning polytrees that are backs of using Bayesian networks is that this inference task close to trees has been considered. More precisely, it is NP-hard. Moreover, the task of constructing a Bayesian has been shown that the best polytree among those that network with an optimal network structure is NP-hard as can be transformed into a tree by at most k edge dele- O(k) O(1) well, even on very restricted instances [Chickering, 1995]. tions can be foundin n |I| time [Gaspers et al., 2015; In this problem, we are given a local parent score func- Safaei et al., 2013] where n is the number of variables and |I| N\{v} is the overall input size. Thus, the running time of these algo- tion fv : 2 → N0 for each variable v and the task is to learn a DAG (N, A) such that the sum of the parent scores rithms is polynomial for every fixed k. As noted by Gaspers over all variables is maximal. et al. [2015], a brute-force algorithm for POLYTREE LEARN- ING would need to consider nn−2 · 2n−1 directed trees. Poly- ∗Supported by the Deutsche Forschungsgemeinschaft (DFG), trees have been used, for example, in image-segmentation for project OPERAH, KO 3669/5-1. microscopy data [Fehri et al., 2019]. A Our Results. We obtain an algorithm that solves POLY- that has no children is a sink.Given a vertex v, we let Pv := TREE LEARNING in 3n · |I|O(1) time. This is the first algo- {u | (u, v) ∈ A} denote the parent set of v. The skeleton rithm for POLYTREE LEARNING where the running time is of a directed graph (N, A) is the undirected graph (N, E) singly-exponential in the number of variables n. Therunning where {u, v} ∈ E if and only if (u, v) ∈ A or (v,u) ∈ time is a substantial improvement over the brute-force algo- A. A directed acyclic graph is a polytree if its skeleton is a rithm mentioned above, thus positively answering a question forest, that is, the skeleton is acyclic [Dasgupta, 1999]. Asa of Gaspers et al. [2015] on the existence of such algorithms. shorthand, we write P × v := P ×{v} for a vertex v and a We then study whether POLYTREE LEARNING is amenable set P ⊆ N. to polynomial-time data reduction that shrinks the instance I Problem Definition. Given a vertex set N, a family F := if |I| is much bigger than n. Using tools from parameter- {f : 2N\{v} → N | v ∈ N} is a family of local scores ized complexity theory, we show that such a data reduction v 0 for N. Intuitively, fv(P ) is the score that a vertex v obtains algorithm is unlikely. More precisely, we show that (un- if P is its parent set. Given a directed graph D := (N, A) we der standard complexity-theoretic assumptions) there is no A define score(A) := v∈N fv(Pv ). We study the following polynomial-time algorithm that replaces any n-variable in- computational problem. stance I by an equivalent one of size nO(1). In other words, P it seems necessary to keep an exponential number of parent POLYTREE LEARNING sets to compute the optimal polytree. Input: A set of vertices N, local scores F = {fv | v ∈ We then consider a parameter that is potentially much N}, and an integer t ∈ N0. Question smaller than n and determine whether POLYTREE LEARN- : Is there an arc-set A ⊆ N × N such ING can be solved efficiently when this parameter is small. that (N, A) is a polytree and score(A) ≥ t? The parameter d, whichwe callthenumberof dependent vari- Given an instance I := (N, F,t) of POLYTREE LEARN- ables, is the number of variables v for which at least one en- ING, an arc-set A is a solution of I if (N, A) is a try of fv is strictly positive. The parameter essentially counts polytree and score(A) ≥ t. Without loss of general- how many variables might receive a nonemptyparent set in an ity we may assume that fv(∅) = 0 for every v ∈ optimal solution. We show that POLYTREE LEARNING can N [Grüttemeier and Komusiewicz, 2020a]. Throughout this be solved in polynomial time when d is constant but that an work, we let n := |N|. O(1) algorithm with running time g(d) · |I| is unlikely for any Solution Structure and Input Representation. We as- computable function g. Consequently, in order to obtain pos- sume that the local scores F are given in non-zero represen- itive results for the parameter d, one needs to consider further tation. That is, f (P ) is part of the input if it is different restrictions on the structure of the input instance. We make a v from 0. For N = {v1,...,vn}, the local scores F are repre- first step in this direction and consider the case where all par- sented by a two-dimensional array [Q ,Q ,...,Q ], where ent sets with a strictly positive score have size at most p. Us- 1 2 n each Qi is an array containing all triples (fvi (P ), |P |, P ) ing this parameterization, we show that every input instance where f (P ) > 0. The size |F| is defined as the number ωdp O(1) vi can be solved in 2 ·|I| time where ω is the matrix mul- of bits needed to store this two-dimensional array. Given an tiplication constant. With the current-best known value for ω instance I := (N, F,t), we define |I| := n + |F| + log(t). dp O(1) this gives a running time of 5.18 · |I| . We then consider For a vertex v, we call PF (v) := {P ⊆ N \ {v} | again data reduction approaches. This time we obtain a posi- fv(P ) > 0} ∪ {∅} the set of potential parents of v. Given tive result: Any instance of POLYTREE LEARNING where p is a yes-instance I := (N, F,t) of POLYTREE LEARNING, constant can be reduced in polynomial time to an equivalent A there exists a solution A such that Pv ∈ PF (v) for ev- O(1) A one of size d . Informally, this means that if the instance ery v ∈ N: If there is a vertex with Pv 6∈ PF (v) has only few dependent variables, the parent sets with strictly we can simply replace its parent set by ∅.
Load more

Recommended publications
  • Context-Bounded Analysis of Concurrent Queue Systems*
    Context-Bounded Analysis of Concurrent Queue Systems Salvatore La Torre1, P. Madhusudan2, and Gennaro Parlato1,2 1 Universit`a degli Studi di Salerno, Italy 2 University of Illinois at Urbana-Champaign, USA Abstract. We show that the bounded context-switching reachability problem for concurrent finite systems communicating using unbounded FIFO queues is decidable, where in each context a process reads from only one queue (but is allowed to write onto all other queues). Our result also holds when individual processes are finite-state recursive programs provided a process dequeues messages only when its local stack is empty. We then proceed to classify architectures that admit a decidable (un- bounded context switching) reachability problem, using the decidability of bounded context switching. We show that the precise class of decid- able architectures for recursive programs are the forest architectures, while the decidable architectures for non-recursive programs are those that do not have an undirected cycle. 1 Introduction Networks of concurrent processes communicating via message queues form a very natural and useful model for several classes of systems with inherent paral- lelism. Two natural classes of systems can be modeled using such a framework: asynchronous programs on a multi-core computer and distributed programs com- municating on a network. In parallel programming languages for multi-core or even single-processor sys- tems (e.g., Java, web service design), asynchronous programming or event-driven programming is a common idiom that programming languages provide [19,7,13]. In order to obtain higher performance and low latency, programs are equipped with the ability to issue tasks using asynchronous calls that immediately return, but are processed later, either in parallel with the calling module or perhaps much later, depending on when processors and other resources such as I/O be- come free.
    [Show full text]
  • Bayesian Networks
    Bayesian networks Charles Elkan October 23, 2012 Important: These lecture notes are based closely on notes written by Lawrence Saul. Text may be copied directly from his notes, or paraphrased. Also, these type- set notes lack illustrations. See the classroom lectures for figures and diagrams. Probability theory is the most widely used framework for dealing with uncer- tainty because it is quantitative and logically consistent. Other frameworks either do not lead to formal, numerical conclusions, or can lead to contradictions. 1 Random variables and graphical models Assume that we have a set of instances and a set of random variables. Think of an instance as being, for example, one patient. A random variable (rv for short) is a measurement. For each instance, each rv has a certain value. The name of an rv is often a capital letter such as X. An rv may have discrete values, or it may be real-valued. Terminology: Instances and rvs are universal in machine learning and statis- tics, so many other names are used for them depending on the context and the application. Generally, an instance is a row in a table of data, while an rv is a column. An rv may also be called a feature or an attribute. It is vital to be clear about the difference between an rv and a possible value of the rv. Values are also called outcomes. One cell in a table of data contains the particular outcome of the column rv for the row instance. If this outcome is unknown, we can think of a special symbol such as “?” being in the cell.
    [Show full text]
  • Lecture 10: Variable Eliminaton, Contnued
    Graphical Models Lecture 10: Variable Eliminaon, con:nued Andrew McCallum [email protected] Thanks to Noah Smith and Carlos Guestrin for some slide materials.1 Last Time • Probabilis:c inference is the Goal: P(X | E = e). – #P-complete in General • Do it anyway! Variable eliminaon … 0 1 Markov Chain Example A P (B)= P (A = a)P (B A = a) P(B |A) 0 1 | a Val(A) ∈ 0 B 1 P (C)= P (B = b)P (C B = b) P(C |B) 0 1 | C b Val(B) 0 ∈ 1 P (D)= P (C = c)P (D C = c) P(D |C) 0 1 D | c Val(C) ∈ 0 1 Last Time • Probabilis:c inference is the Goal: P(X | E = e). – #P-complete in General • Do it anyway! Variable eliminaon … – Work on factors (alGebra of factors) – Generally: “sum-product” inference φ Z φ Φ ∈ Products of Factors • Given two factors with different scopes, we can calculate a new factor equal to their products. A B C ϕ3(A, B, C) 0 0 0 3000 0 0 1 30 0 1 0 5 A B ϕ1(A, B) B C ϕ2(B, C) 0 0 30 0 0 100 0 1 1 500 0 1 5 0 1 1 = 1 0 0 100 1 0 1 . 1 0 1 1 0 1 1 1 1 10 1 1 100 1 1 0 10 1 1 1 1000 Factor MarGinalizaon • Given X and Y (Y ∉ X), we can turn a factor ϕ(X, Y) into a factor ψ(X) via marGinalizaon: ψ(X)= φ(X,y) y Val(Y ) ∈ P(C | A, B) 0, 0 0, 1 1, 0 1,1 A C ψ(A, C) 0 0.5 0.4 0.2 0.1 0 0 0.9 1 0.5 0.6 0.8 0.9 0 1 0.3 1 0 1.1 “summinG out” B 1 1 1.7 Last Time • Probabilis:c inference is the Goal: P(X | E = e).
    [Show full text]
  • Polytree-Augmented Classifier Chains for Multi-Label Classification
    Proceedings of the Twenty-Fourth International Joint Conference on Artificial Intelligence (IJCAI 2015) Polytree-Augmented Classifier Chains for Multi-Label Classification Lu Sun and Mineichi Kudo Graduate School of Information Science and Technology Hokkaido University, Sapporo, Japan fsunlu, [email protected] Abstract Multi-label classification is a challenging and ap- pealing supervised learning problem where a sub- set of labels, rather than a single label seen in tra- ditional classification problems, is assigned to a single test instance. Classifier chains based meth- ods are a promising strategy to tackle multi-label classification problems as they model label corre- (a) “fish”,“sea” (b) “windsock”,“sky” lations at acceptable complexity. However, these methods are difficult to approximate the under- Figure 1: A multi-label example to show label dependency lying dependency in the label space, and suffer for inference. (a) A fish in the sea; (b) carp shaped windsocks from the problems of poorly ordered chain and (koinobori) in the sky. error propagation. In this paper, we propose a novel polytree-augmented classifier chains method to remedy these problems. A polytree is used contexts (backgrounds), in other words, the label dependen- to model reasonable conditional dependence be- cies, distinguish them clearly. tween labels over attributes, under which the di- The existing MLC methods can be cast into two rectional relationship between labels within causal strategies: problem transformation and algorithm adapta- basins could be appropriately determined. In ad- tion [Tsoumakas and Katakis, 2007]. A convenient and dition, based on the max-sum algorithm, exact in- straightforward way for MLC is to conduct problem trans- ference would be performed on polytrees at rea- formation in which a MLC problem is transformed into sonable cost, preventing from error propagation.
    [Show full text]
  • Arxiv:1802.06316V1 [Math.AC] 18 Feb 2018 Ne Yl,Eg Ideal
    PROJECTIVE DIMENSION AND REGULARITY OF EDGE IDEAL OF SOME WEIGHTED ORIENTED GRAPHS ∗ GUANGJUN ZHU , LI XU, HONG WANG AND ZHONGMING TANG Abstract. In this paper we provide some exact formulas for the projective di- mension and the regularity of edge ideals associated to vertex weighted rooted forests and oriented cycles. As some consequences, we give some exact formulas for the depth of these ideals. 1. Introduction A directed graph D consists of a finite set V (D) of vertices, together with a prescribed collection E(D) of ordered pairs of distinct points called edges or arrows. If {u, v} ∈ E(D) is an edge, we write uv for {u, v}, which is denoted to be the directed edge where the direction is from u to v and u (resp. v) is called the starting point (resp. the ending point). An oriented graph is a directed graph having no bidirected edges (i.e. each pair of vertices is joined by a single edge having a unique direction). In other words an oriented graph D is a simple graph G together with an orientation of its edges. We call G the underlying graph of D. An oriented graph having no multiple edges (edges with same starting point and ending point) or loops (edges that connect vertices to themselves) is called a simple oriented graph. An oriented cycle is a cycle graph in which all edges are oriented in clockwise or in counterclockwise. An oriented acyclic graph is a simple directed graph without oriented cycles. An oriented tree or polytree is a oriented acyclic graph formed by orienting the edges of undirected acyclic graphs.
    [Show full text]
  • Introduction to Neurabase: Forming Join Trees in Bayesian Networks
    Introduction to NeuraBASE: Forming Join Trees in Bayesian Networks Robert Hercus Choong-Ming Chin Kim-Fong Ho Introduction A Bayesian network or a belief network is a condensed graphical model of probabilistic relationships amongst a set of variables and is also a popular representation to encode uncertainties in expert systems. In general, Bayesian networks can be designed and trained to provide future classifications using a set of attributes derived from statistical data analysis. Other methods for learning Bayesian networks from data have also been developed (see Heckerman [2]). In some cases, they can also handle situations where data entries are missing. Bayesian networks are often associated with the notion of causality and for a network to be considered a Bayesian network, the following requirements (see Jensen [3]) must hold: ● A set of random variables and a set of directed edges between variables must exist. ● Each variable must have a finite set of mutually exclusive states. ● The variables together with the directed edges must form a directed acyclic graph (DAG). (A directed graph is acyclic if there is no directed path A1 → . An such that A1 = An). ● For each variable A with parents B1, B2, . , Bn there is an attached potentiated table P(A|B1, B2, . , Bn). In this paper, its main objective is to provide an alternative method to find clique trees that are paramount for the efficient computation of probabilistic enquiries posed to Bayesian networks. For a simple Bayesian network, as shown in Figure 1, given the evidence that people are leaving, we may want to know what is the probability that fire has occurred, or even the probability that there is smoke, given that the alarm has gone off.
    [Show full text]
  • The Study of the Theoretical Size and Node Probability of the Loop Cutset in Bayesian Networks
    mathematics Article The Study of the Theoretical Size and Node Probability of the Loop Cutset in Bayesian Networks Jie Wei , Yufeng Nie * and Wenxian Xie School of Mathematics and Statistics, Northwestern Polytechnical University, Xi’an 710129, China; [email protected] (J.W.); [email protected] (W.X.) * Correspondence: [email protected] Received: 27 May 2020; Accepted: 1 July 2020; Published: 3 July 2020 Abstract: Pearl’s conditioning method is one of the basic algorithms of Bayesian inference, and the loop cutset is crucial for the implementation of conditioning. There are many numerical algorithms for solving the loop cutset, but theoretical research on the characteristics of the loop cutset is lacking. In this paper, theoretical insights into the size and node probability of the loop cutset are obtained based on graph theory and probability theory. It is proven that when the loop cutset in a p-complete graph has a size of p − 2, the upper bound of the size can be determined by the number of nodes. Furthermore, the probability that a node belongs to the loop cutset is proven to be positively correlated with its degree. Numerical simulations show that the application of the theoretical results can facilitate the prediction and verification of the loop cutset problem. This work is helpful in evaluating the performance of Bayesian networks. Keywords: Bayesian inference; loop cutset; node probability of loop cutset; conditioning method 1. Introduction A Bayesian network is an acyclic, directed graph in which the nodes represent chance variables. The arcs in a Bayesian network represent the probabilistic influences between nodes.
    [Show full text]
  • Learning Polytrees
    Learning Polytrees Sanjoy Dasgupta Department of Computer Science University of California, Berkeley Abstract NP-hardness result closes one door but opens many oth- ers. Over the last two decades, provably good approxi- mation algorithms have been developed for a wide host of We consider the task of learning the maximum- NP-hard optimization problems, including for instance the likelihood polytree from data. Our first result is a Euclidean travelling salesman problem. Are there approxi- performance guarantee establishing that the opti- mation algorithms for learning structure? mal branching (or Chow-Liu tree), which can be computed very easily, constitutes a good approx- We take the first step in this direction by considering the imation to the best polytree. We then show that task of learning polytrees from data. A polytree is a di- it is not possible to do very much better, since rected acyclic graph with the property that ignoring the di- the learning problem is NP-hard even to approx- rections on edges yields a graph with no undirected cycles. imately solve within some constant factor. Polytrees have more expressive power than branchings, and have turned out to be a very important class of directed probabilistic nets, largely because they permit fast exact in- 1 Introduction ference (Pearl, 1988). In the setup we will consider, we are given i.i.d. samples from an unknown distribution D over (X1,...,X ) and we wish to find some polytree with n The huge popularity of probabilistic nets has created a n nodes, one per X , which models the data well. There are pressing need for good algorithms that learn these struc- i no assumptions about D.
    [Show full text]
  • An Algorithm to Learn Polytree Networks with Hidden Nodes
    An Algorithm to Learn Polytree Networks with Hidden Nodes Firoozeh Sepehr Donatello Materassi Department of EECS Department of EECS University of Tennessee Knoxville University of Tennessee Knoxville 1520 Middle Dr, Knoxville, TN 37996 1520 Middle Dr, Knoxville, TN 37996 [email protected] [email protected] Abstract Ancestral graphs are a prevalent mathematical tool to take into account latent (hid- den) variables in a probabilistic graphical model. In ancestral graph representations, the nodes are only the observed (manifest) variables and the notion of m-separation fully characterizes the conditional independence relations among such variables, bypassing the need to explicitly consider latent variables. However, ancestral graph models do not necessarily represent the actual causal structure of the model, and do not contain information about, for example, the precise number and location of the hidden variables. Being able to detect the presence of latent variables while also inferring their precise location within the actual causal structure model is a more challenging task that provides more information about the actual causal rela- tionships among all the model variables, including the latent ones. In this article, we develop an algorithm to exactly recover graphical models of random variables with underlying polytree structures when the latent nodes satisfy specific degree conditions. Therefore, this article proposes an approach for the full identification of hidden variables in a polytree. We also show that the algorithm is complete in the sense that when such degree conditions are not met, there exists another polytree with fewer number of latent nodes satisfying the degree conditions and entailing the same independence relations among the observed variables, making it indistinguishable from the actual polytree.
    [Show full text]
  • Arxiv:1507.02608V6 [Math.ST] 3 Feb 2018 Represented by a Completed Partially Directed Acyclic Graph (CPDAG)
    Submitted to the Annals of Statistics HIGH-DIMENSIONAL CONSISTENCY IN SCORE-BASED AND HYBRID STRUCTURE LEARNING By Preetam Nandyy,∗, Alain Hauserz and Marloes H. Maathuisy,∗ ETH Z¨urichy and Bern University of Applied Sciencesz Main approaches for learning Bayesian networks can be classified as constraint-based, score-based or hybrid methods. Although high- dimensional consistency results are available for constraint-based meth- ods like the PC algorithm, such results have not been proved for score- based or hybrid methods, and most of the hybrid methods have not even shown to be consistent in the classical setting where the num- ber of variables remains fixed and the sample size tends to infinity. In this paper, we show that consistency of hybrid methods based on greedy equivalence search (GES) can be achieved in the classical set- ting with adaptive restrictions on the search space that depend on the current state of the algorithm. Moreover, we prove consistency of GES and adaptively restricted GES (ARGES) in several sparse high-dimensional settings. ARGES scales well to sparse graphs with thousands of variables and our simulation study indicates that both GES and ARGES generally outperform the PC algorithm. 1. Introduction. A Bayesian network consists of a directed acyclic graph (DAG) on a set of variables and conditional distributions for each node given its parents in the DAG. Bayesian networks can be used for various purposes, such as efficiently modeling the joint distribution of the variables, constructing decision support systems, probabilistic reasoning in expert sys- tems, and causal inference. In a Bayesian network, the DAG encodes conditional independence rela- tionships that must hold among the corresponding random variables.
    [Show full text]
  • Graph Theory
    Definitions Graph Theory Types of Graphs Trajectories and Circuits Probabilistic Graphical Models Graph Isomorphism L. Enrique Sucar, INAOE Trees Cliques Perfect Ordering Ordering and Triangulation Algorithms References (INAOE) 1 / 32 Outline Definitions 1 Definitions Types of Graphs 2 Types of Graphs Trajectories and Circuits Graph 3 Trajectories and Circuits Isomorphism Trees 4 Graph Isomorphism Cliques Perfect 5 Trees Ordering Ordering and 6 Cliques Triangulation Algorithms References 7 Perfect Ordering 8 Ordering and Triangulation Algorithms 9 References (INAOE) 2 / 32 Definitions Graphs Definitions Types of • A graph provides a compact way to represent binary Graphs relations between a set of objects Trajectories and Circuits • Objects are represented as circles or ovals, and Graph Isomorphism relations as lines or arrows Trees • There are two basic types of graphs: undirected graphs Cliques and directed graphs Perfect Ordering Ordering and Triangulation Algorithms References (INAOE) 3 / 32 Definitions Directed Graphs Definitions Types of Graphs Trajectories and Circuits Graph Isomorphism • A directed graph or digraph is an ordered pair, Trees G = (V ; E), where V is a set of vertices or nodes and E Cliques is a set of arcs that represent a binary relation on V Perfect Ordering • Directed graphs represent anti-symmetric relations Ordering and between objects, for instance the “parent” relation Triangulation Algorithms References (INAOE) 4 / 32 Definitions Undirected Graphs Definitions Types of Graphs Trajectories and Circuits Graph • An undirected
    [Show full text]
  • Inference in Bayesian Networks
    3 Inference in Bayesian Networks 3.1 Introduction The basic task for any probabilistic inference system is to compute the posterior probability distribution for a set of query nodes, given values for some evidence nodes. This task is called belief updating or probabilistic inference. Inference in Bayesian networks is very flexible, as evidence can be entered about any node while beliefs in any other nodes are updated. In this chapter we will cover the major classes of inference algorithms — exact and approximate — that have been developed over the past 20 years. As we will see, different algorithms are suited to different network structures and performance requirements. Networks that are simple chains merely require repeated application of Bayes’ Theorem. Inference in simple tree structures can be done using local computations and message passing between nodes. When pairs of nodes in the BN are connected by multiple paths the inference algorithms become more complex. For some networks, exact inference becomes computationally infeasible, in which case approximate inference algorithms must be used. In general, both exact and approximate inference are theoretically computationally complex (specifically, NP hard). In practice, the speed of inference depends on factors such as the structure of the network, including how highly connected it is, how many undirected loops there are and the locations of evidence and query nodes. Many inference algorithms have not seen the light of day beyond the research environment that produced them. Good exact and approximate inference algorithms are implemented in BN software, so knowledge engineers do not have to. Hence, our main focus is to characterize the main algorithms’ computational performance to both enhance understanding of BN modeling and help the knowledge engineer assess which algorithm is best suited to the application.
    [Show full text]