Recent Development on Elimination Ordering Group 1.

Slides:



Advertisements
Similar presentations
Minimum Clique Partition Problem with Constrained Weight for Interval Graphs Jianping Li Department of Mathematics Yunnan University Jointed by M.X. Chen.
Advertisements

Variational Methods for Graphical Models Micheal I. Jordan Zoubin Ghahramani Tommi S. Jaakkola Lawrence K. Saul Presented by: Afsaneh Shirazi.
Lauritzen-Spiegelhalter Algorithm
1 EE5900 Advanced Embedded System For Smart Infrastructure Static Scheduling.
Exact Inference in Bayes Nets
“Devo verificare un’equivalenza polinomiale…Che fò? Fò dù conti” (Prof. G. Di Battista)
Dynamic Programming.
Junction Trees And Belief Propagation. Junction Trees: Motivation What if we want to compute all marginals, not just one? Doing variable elimination for.
1 NP-Complete Problems. 2 We discuss some hard problems:  how hard? (computational complexity)  what makes them hard?  any solutions? Definitions 
Dynamic Bayesian Networks (DBNs)
Combinatorial Algorithms
Effective Heuristics for NP-Hard Problems Arising in Molecular Biology Richard M. Karp Bangalore, January 5, 2011.
Junction Trees: Motivation Standard algorithms (e.g., variable elimination) are inefficient if the undirected graph underlying the Bayes Net contains cycles.
GS 540 week 6. HMM basics Given a sequence, and state parameters: – Each possible path through the states has a certain probability of emitting the sequence.
02/01/11CMPUT 671 Lecture 11 CMPUT 671 Hard Problems Winter 2002 Joseph Culberson Home Page.
Clustering short time series gene expression data Jason Ernst, Gerard J. Nau and Ziv Bar-Joseph BIOINFORMATICS, vol
Graph Algorithms: Minimum Spanning Tree We are given a weighted, undirected graph G = (V, E), with weight function w:
3 -1 Chapter 3 The Greedy Method 3 -2 The greedy method Suppose that a problem can be solved by a sequence of decisions. The greedy method has that each.
1 Optimization problems such as MAXSAT, MIN NODE COVER, MAX INDEPENDENT SET, MAX CLIQUE, MIN SET COVER, TSP, KNAPSACK, BINPACKING do not have a polynomial.
MAE 552 – Heuristic Optimization Lecture 6 February 6, 2002.
Implicit Hitting Set Problems Richard M. Karp Harvard University August 29, 2011.
NP-Complete Problems Reading Material: Chapter 10 Sections 1, 2, 3, and 4 only.
Graph Triangulation by Dmitry Pidan Based on the paper “A sufficiently fast algorithm for finding close to optimal junction tree” by Ann Becker and Dan.
Randomness in Computation and Communication Part 1: Randomized algorithms Lap Chi Lau CSE CUHK.
Ryan Kinworthy 2/26/20031 Chapter 7- Local Search part 2 Ryan Kinworthy CSCE Advanced Constraint Processing.
CSE 589 Applied Algorithms Spring Colorability Branch and Bound.
1 Introduction to Approximation Algorithms. 2 NP-completeness Do your best then.
© The McGraw-Hill Companies, Inc., Chapter 3 The Greedy Method.
Fixed Parameter Complexity Algorithms and Networks.
Stochastic Algorithms Some of the fastest known algorithms for certain tasks rely on chance Stochastic/Randomized Algorithms Two common variations – Monte.
Approximating Minimum Bounded Degree Spanning Tree (MBDST) Mohit Singh and Lap Chi Lau “Approximating Minimum Bounded DegreeApproximating Minimum Bounded.
1 Introduction to Approximation Algorithms. 2 NP-completeness Do your best then.
Tonga Institute of Higher Education Design and Analysis of Algorithms IT 254 Lecture 8: Complexity Theory.
1 Simulated Annealing Contents 1. Basic Concepts 2. Algorithm 3. Practical considerations.
Chapter 15 Approximation Algorithm Introduction Basic Definition Difference Bounds Relative Performance Bounds Polynomial approximation Schemes Fully Polynomial.
1 Variable Elimination Graphical Models – Carlos Guestrin Carnegie Mellon University October 11 th, 2006 Readings: K&F: 8.1, 8.2, 8.3,
CSE332: Data Abstractions Lecture 24.5: Interlude on Intractability Dan Grossman Spring 2012.
Inference Complexity As Learning Bias Daniel Lowd Dept. of Computer and Information Science University of Oregon Joint work with Pedro Domingos.
Course: Logic Programming and Constraints
Learning With Bayesian Networks Markus Kalisch ETH Zürich.
CSE 589 Part VI. Reading Skiena, Sections 5.5 and 6.8 CLR, chapter 37.
NP-Complete Problems. Running Time v.s. Input Size Concern with problems whose complexity may be described by exponential functions. Tractable problems.
NP-COMPLETE PROBLEMS. Admin  Two more assignments…  No office hours on tomorrow.
Approximate Inference: Decomposition Methods with Applications to Computer Vision Kyomin Jung ( KAIST ) Joint work with Pushmeet Kohli (Microsoft Research)
CS270 Project Overview Maximum Planar Subgraph Danyel Fisher Jason Hong Greg Lawrence Jimmy Lin.
Implicit Hitting Set Problems Richard M. Karp Erick Moreno Centeno DIMACS 20 th Anniversary.
Optimization Problems
Exact Inference in Bayes Nets. Notation U: set of nodes in a graph X i : random variable associated with node i π i : parents of node i Joint probability:
Algorithms for hard problems Introduction Juris Viksna, 2015.
Today Graphical Models Representing conditional dependence graphically
1 Structure Learning (The Good), The Bad, The Ugly Inference Graphical Models – Carlos Guestrin Carnegie Mellon University October 13 th, 2008 Readings:
1 Variable Elimination Graphical Models – Carlos Guestrin Carnegie Mellon University October 15 th, 2008 Readings: K&F: 8.1, 8.2, 8.3,
The NP class. NP-completeness Lecture2. The NP-class The NP class is a class that contains all the problems that can be decided by a Non-Deterministic.
Unconstrained Submodular Maximization Moran Feldman The Open University of Israel Based On Maximizing Non-monotone Submodular Functions. Uriel Feige, Vahab.
Optimization Problems
Independent Cascade Model and Linear Threshold Model
Inference in Bayesian Networks
Approximation algorithms
Algorithms for hard problems
Exact Inference Continued
Independent Cascade Model and Linear Threshold Model
Randomized Algorithms CS648
Optimization Problems
CSE 589 Applied Algorithms Spring 1999
Richard Anderson Lecture 29 Complexity Theory
NP-Complete Problems.
Exact Inference Continued
Variable Elimination Graphical Models – Carlos Guestrin
Independent Cascade Model and Linear Threshold Model
Presentation transcript:

Recent Development on Elimination Ordering Group 1

Recent Development on Elimination Ordering By Florence Lo Andrew Yan

Table of Contents Problem Statement Background Motivations Approaches to Elimination Orderings An approximation algorithm for triangulation Q & A

Problem Statement Minimizing Total Cost in Vertices Elimination

Background BN specifies a complete joint probability distribution (JPD) over all variables. Given the JPD, one can answer all possible inference queries by summing out over irrelevant variables. Assumed all variables have 2 states and there are n variables => O(2 n )

Background (cont) To make inference more efficient, 2 exact inference algorithms are used 1) Variable Elimination Algorithm (VE) 2) Clique Tree Propagation (CTP)

Variable Elimination (VE) Use the factored representation of the JPD to do marginalization efficiently. The key idea is to “push sum in” as far as possible when summing out irrelevant terms Complexity of VE is based on the cost of variables elimination (i.e. elimination orderings)

Clique Tree Propagation (CTP) Inference in BN is formulated as message passing in a junction tree First step in CTP is triangulated graph, one way is to eliminate vertices one by one to add the extra edges (i.e. elimination orderings)

Discussion Different elimination ordering leads to different cost NP-hard

3 Approaches to Establish Elimination Orderings 1. Elimination Ordering Heuristics 2. Triangulation 3. Simulated Annealing

Elimination Ordering Heuristics Maximum Cardinality Search (Tarjan et. al. 1985) Minimum Deficiency Heuristics (Bertele et. al. 1972) Minimum Degree Heuristics (Rose 1972)

Discussions Easy to implement Concepts are easy Linear time complexity Good approximation to the optimal solution

Triangulation Objectives 1. Size of the maximal clique is minimum 2. Minimal triangulation

Minimal Triangulation Computing a minimal triangulation consists in embedding a given graph with a triangulated graph by adding a set of edges (called a fill). If the set of edges added is inclusion-minimal, the fill is said to be minimal, and the corresponding triangulated graph is called a minimal triangulation

Discussion Minimal triangulation generally improves the total cost May stuck in the local minima of the cost function NP-hard

Simulated Annealing Stochastic optimization algorithm to find global minimum cost configuration of NP- complete combinatorial problems with cost functions having many local minima A combination of deterministic descent search and a Monte Carlo method Accept cost function increase with positive probability dependent on the state of the search process

Pseudo -code Select an initial solution s 0 Select an initial temperature t 0 > 0 Select a temperature reduction function α Repeat repeat randomly select s from N(s 0 ) δ = f(s) – f(s 0 ) If δ < 0 then s 0 = s else generate random x uniformly in the range (0, 1) if x < exp(- δ/t) then s 0 = s until iteration_count = nrep Set t = α(t) Until stopping condition = true S 0 is the approximation to the optimal solution

2 Categories of Decisions to make 1. Parameters of the annealing algorithm - t0, cooling schedule governed by nrep and α, stopping conditions 2. Problem specific and involve the choice of the space of feasible solutions, the form of the cost function, neighborhood structure employed

Discussion Time consuming Extensive experiments on the control parameters Enhancement and modifications to speed up computation time

An approximation algorithm for triangulation By Ann Becker and Dan Geiger in 1996 Same state space size Optimality criterion – Cliquewidth, k Cliquewidth 2  +1 – approximation –  = approximation ratio for 3-way vertex cut3-way vertex cut O(2 (2  +1)k npoly(n)) – poly(n) = linear programming Divide and conquer

The algorithm Triangulate(G,W,k) If |V| < (2  +1)K then make a clique out of G Else Find a W-decomposition (X,A,B,C) of G wrt (k,  )W-decomposition If not found Return “cliquewidth > k” W A  W  A, W B  W  B, W C  W  C Call Triangulate( G[A  X], W A,k) Call Triangulate( G[B  X], W B,k) Call Triangulate( G[C  X], W C,k) Make a clique of G[W  X]

Trial and error Try for k = 1, 2, 3… until success!!!

Example (k=3,  =1)

Improvement Processing the input – Simplicial vertex – Removed repeatedly – Improves the running time

Improvement Termination condition – |V| < (2  +1)k  from a clique – Junction tree instead – W as a clique – Approximation may improved

Improvement Post-processing the output – May not minimal triangulated – Remove some added edges, still triangulated – Kjaerulff’s algorithm

In practice Time complexity O(2 (2  +1)k npoly(n)) |W| < k in most case – k  2 2k W consists of two subsets – No 3-way vertex cut Step further for k Find a W-decomposition (X,A,B,C) of G wrt (k,  ) If not found Return “cliquewidth > k”

A tighter bound l be the size of the largest clique in the output Test against (2  +1)k – The ratio can be smaller than 2  +1 – Optimal cliquewidth can be larger than k l/k instead of 2  +1 Instance-specific posteriori bound

The weighted problem Different state spaces of each vertex – w(v) = log 2 (state space size) – w(clique) = sum of w(v) in the clique Weighted W-decomposition When terminate in recursion, run greedy algorithm (minimum weight heuristics) (2  +1)m

Results Medianus I (43 vertices, 110 edges) Compare with the enhanced minimum weight heuristics – Better when state space increase l/k = 10/6 (not 3.66) Run time: one or two minutes

Discussion O(2 4.66k npoly(n)) – Polynomial for k = O(log n) Exponential time for an arbitrary inference

What is cliquewidth? Undirected graph The size of the largest clique in the junction free of the graph in which the size of the largest clique is minimized back

What is 3-way vertex cut? A weighted undirected graph 3 vertices A set of vertices of minimum weight whose removal  three vertices disconnected 4/3 – approximation 2 – approximation back

What is W-decomposition? An integer k  1, a real number   1 A graph G = ( V, E ), |V|  ( 2  + 1 ) k W  V A decomposition ( X, A, B, C ) wrt ( k,  )decomposition – |W| < (  + 1 )k – |X| <  k – |(W  A)  X| < (  + 1 )k – |(W  B)  X| < (  + 1 )k – |(W  C)  X| < (  + 1 )k back

What is decomposition? A graph G = ( V, E ) A partition ( X, A, B, C ) – A, B   – No edges between A, B, C back