Junction Tree Algorithm 1. David Barber

Similar documents
From Elimination to Belief Propagation

Why the Junction Tree Algorithm? The Junction Tree Algorithm. Clique Potential Representation. Overview. Chris Williams 1.

Assignment 4 Biophys 4322/5322

Utilizing exact and Monte Carlo methods to investigate properties of the Blume Capel Model applied to a nine site lattice.

Abstract Interpretation: concrete and abstract semantics

Basic Polyhedral theory

Elements of Statistical Thermodynamics

MA1506 Tutorial 2 Solutions. Question 1. (1a) 1 ) y x. e x. 1 exp (in general, Integrating factor is. ye dx. So ) (1b) e e. e c.

AP Calculus BC Problem Drill 16: Indeterminate Forms, L Hopital s Rule, & Improper Intergals

( ) Differential Equations. Unit-7. Exact Differential Equations: M d x + N d y = 0. Verify the condition

Strongly Connected Components

VTU NOTES QUESTION PAPERS NEWS RESULTS FORUMS

cycle that does not cross any edges (including its own), then it has at least

Linked-List Implementation. Linked-lists for two sets. Multiple Operations. UNION Implementation. An Application of Disjoint-Set 1/9/2014

The graph of y = x (or y = ) consists of two branches, As x 0, y + ; as x 0, y +. x = 0 is the

Hydrogen Atom and One Electron Ions

Thus, because if either [G : H] or [H : K] is infinite, then [G : K] is infinite, then [G : K] = [G : H][H : K] for all infinite cases.

Addition of angular momentum

Chemical Physics II. More Stat. Thermo Kinetics Protein Folding...

Exam 1. It is important that you clearly show your work and mark the final answer clearly, closed book, closed notes, no calculator.

1 Minimum Cut Problem

Week 3: Connected Subgraphs

Fourier Transforms and the Wave Equation. Key Mathematics: More Fourier transform theory, especially as applied to solving the wave equation.

Propositional Logic. Combinatorial Problem Solving (CPS) Albert Oliveras Enric Rodríguez-Carbonell. May 17, 2018

Complex Powers and Logs (5A) Young Won Lim 10/17/13

The Matrix Exponential

International Journal of Scientific & Engineering Research, Volume 6, Issue 7, July ISSN

Roadmap. XML Indexing. DataGuide example. DataGuides. Strong DataGuides. Multiple DataGuides for same data. CPS Topics in Database Systems

Problem 22: Journey to the Center of the Earth

1 Isoparametric Concept

Modeling Causal Reinforcement And Undermining With Noisy-AND Trees

Network Congestion Games

10. The Discrete-Time Fourier Transform (DTFT)

Brief Introduction to Statistical Mechanics

Numerical methods, Mixed exercise 10

Lecture 37 (Schrödinger Equation) Physics Spring 2018 Douglas Fields

Lie Groups HW7. Wang Shuai. November 2015

Higher order derivatives

A Propagating Wave Packet Group Velocity Dispersion

EXST Regression Techniques Page 1

Derivation of Eigenvalue Matrix Equations

COMPUTER GENERATED HOLOGRAMS Optical Sciences 627 W.J. Dallas (Monday, April 04, 2005, 8:35 AM) PART I: CHAPTER TWO COMB MATH.

The van der Waals interaction 1 D. E. Soper 2 University of Oregon 20 April 2012

First derivative analysis

Background: We have discussed the PIB, HO, and the energy of the RR model. In this chapter, the H-atom, and atomic orbitals.

ANALYSIS IN THE FREQUENCY DOMAIN

Examples and applications on SSSP and MST

Ch. 24 Molecular Reaction Dynamics 1. Collision Theory

Einstein Equations for Tetrad Fields

CS553 Lecture Register Allocation I 3

ECE 407 Computer Aided Design for Electronic Systems. Instructor: Maria K. Michael. Overview. CAD tools for multi-level logic synthesis:

EEO 401 Digital Signal Processing Prof. Mark Fowler

Coupled Pendulums. Two normal modes.

The Equitable Dominating Graph

Derangements and Applications

COMP108 Algorithmic Foundations

11/13/17. directed graphs. CS 220: Discrete Structures and their Applications. relations and directed graphs; transitive closure zybooks

The Matrix Exponential

UNTYPED LAMBDA CALCULUS (II)

As the matrix of operator B is Hermitian so its eigenvalues must be real. It only remains to diagonalize the minor M 11 of matrix B.

PHA 5127 Answers Homework 2 Fall 2001

MCB137: Physical Biology of the Cell Spring 2017 Homework 6: Ligand binding and the MWC model of allostery (Due 3/23/17)

(Upside-Down o Direct Rotation) β - Numbers

Combinatorial Networks Week 1, March 11-12

Computing and Communications -- Network Coding

priority queue ADT heaps 1

y = 2xe x + x 2 e x at (0, 3). solution: Since y is implicitly related to x we have to use implicit differentiation: 3 6y = 0 y = 1 2 x ln(b) ln(b)

Title: Vibrational structure of electronic transition

That is, we start with a general matrix: And end with a simpler matrix:

u 3 = u 3 (x 1, x 2, x 3 )

Supplementary Materials

Final Exam Solutions

Where k is either given or determined from the data and c is an arbitrary constant.

Objective Mathematics

COHORT MBA. Exponential function. MATH review (part2) by Lucian Mitroiu. The LOG and EXP functions. Properties: e e. lim.

a 1and x is any real number.

Probability and Stochastic Processes: A Friendly Introduction for Electrical and Computer Engineers Roy D. Yates and David J.

22/ Breakdown of the Born-Oppenheimer approximation. Selection rules for rotational-vibrational transitions. P, R branches.

ME 321 Kinematics and Dynamics of Machines S. Lambert Winter 2002

Answer Homework 5 PHA5127 Fall 1999 Jeff Stark

EECE 301 Signals & Systems Prof. Mark Fowler

INC 693, 481 Dynamics System and Modelling: Linear Graph Modeling II Dr.-Ing. Sudchai Boonto Assistant Professor

MCE503: Modeling and Simulation of Mechatronic Systems Discussion on Bond Graph Sign Conventions for Electrical Systems

Codes on Planar Graphs

Spectral Synthesis in the Heisenberg Group

Approximate Maximum Flow in Undirected Networks by Christiano, Kelner, Madry, Spielmann, Teng (STOC 2011)

Direct Approach for Discrete Systems One-Dimensional Elements

Chapter 6 Folding. Folding

Search sequence databases 3 10/25/2016

DIFFERENTIAL EQUATION

1997 AP Calculus AB: Section I, Part A

Part 7: Capacitance And Capacitors

Y 0. Standing Wave Interference between the incident & reflected waves Standing wave. A string with one end fixed on a wall

Middle East Technical University Department of Mechanical Engineering ME 413 Introduction to Finite Element Analysis

The second condition says that a node α of the tree has exactly n children if the arity of its label is n.

SCHUR S THEOREM REU SUMMER 2005

First order differential equation Linear equation; Method of integrating factors

Instructions for Section 1

Chapter 8: Electron Configurations and Periodicity

General Notes About 2007 AP Physics Scoring Guidelines

Transcription:

Juntion Tr Algorithm 1 David Barbr Univrsity Collg London 1 Ths slids aompany th book Baysian Rasoning and Mahin Larning. Th book and dmos an b downloadd from www.s.ul.a.uk/staff/d.barbr/brml. Fdbak and orrtions ar also availabl on th sit. Fl fr to adapt ths slids for your own purposs, but plas inlud a link th abov wbsit.

A gnral purpos infrn algorithm (?) Appliability Effiiny Th JTA dals with marginal infrn in multiply-onntd struturs. Th JTA an handl both Blif and Markov Ntworks. Th omplxity of th JTA an b vry high if th graph is multiply onntd. Provids an uppr bound on th omputational omplxity. May b that thr ar som problms for whih muh mor ffiint algorithms xist than th JTA.

Cliqu Graph A liqu graph onsists of a st of potntials, φ 1 (X 1 ),..., φ n (X n ) ah dfind on a st of variabls X i. For nighbouring liqus on th graph, dfind on sts of variabls X i and X j, th intrstion X s = X i X j is alld th sparator and has a orrsponding potntial φ s (X s ). A liqu graph rprsnts th funtion φ (X ) s φ s(x s ) Exampl φ(x 1 )φ(x 2 ) φ(x 1 X 2 ) X 1 X 1 X 2 X 2

Markov Nt Cliqu Graph p(a, b,, d) = φ(a, b, )φ(b,, d) Z b a d (a) a, b, b, b,, d (b) Figur : (a) Markov ntwork φ(a, b, )φ(b,, d). (b) Cliqu graph rprsntation of (a). Cliqu potntial assignmnts Th sparator potntial may b st to th normalisation onstant Z. Cliqus hav potntials φ(a, b, ) and φ(b,, d).

Transformation φ(a, b, )φ(b,, d) p(a, b,, d) = Z By summing w hav Zp(a, b, ) = φ(a, b, ) d φ(b,, d), Zp(b,, d) = φ(b,, d) a φ(a, b, ) Multiplying th two xprssions, ( w hav Z 2 p(a, b, )p(b,, d) = φ(a, b, ) d φ(b,, d) ) ( φ(b,, d) a φ(a, b, ) ) In othr words p(a, b,, d) = = Z 2 p(a, b,, d) a,d p(a, b, )p(b,, d) p(, b) p(a, b,, d) Cliqu potntial assignmnts Th sparator potntial may b st to p(b, ). Cliqus hav potntials p(a, b, ) and p(b,, d). Th liqus and sparators ontain th marginal distributions.

Markov Cliqu Graph Th transformation φ(a, b, ) p(a, b, ) φ(b,, d) p(b,, d) Z p(, b) Th usfulnss of this rprsntation is that if w ar intrstd in th marginal p(a, b, ), this an b rad off from th transformd liqu potntial. JTA Th JTA is a systmati way of transforming th liqu graph potntials so that at th nd of th transformation th nw potntials ontain th marginals of th distribution. Th JTA will work by a squn of loal transformations Eah loal transformation will lav th Cliqu rprsntation invariant.

Absorption Considr nighbouring liqus V and W, sharing th variabls S in ommon. In this as, th distribution on th variabls X = V W is φ (V)φ (W) p(x ) = and our aim is to find a nw rprsntation φ (V) φ (W) p(x ) = ˆφ (V) ˆφ (W) ˆφ (V) ˆ ˆφ (W) ˆ in whih th potntials ar givn by ˆφ (V) = p(v), ˆφ (W) = p(w), ˆ = p(s) W an xpliitly work out th nw potntials as funtion of th old potntials: p(w) = p(x ) = φ (V)φ (W) V\S = φ (W) φ (V) V\S V\S and p(v) = W\S p(x ) = W\S φ (V)φ (W) = φ (V) W\S φ (W)

Absorption W say that th lustr W absorbs information from lustr V. First w dfin a nw sparator φ (S) = V\S φ (V) and rfin th W potntial using φ (V) φ (S) φ (W) φ (W) = φ (W) φ (S) Invarian Th advantag of this intrprtation is that th nw rprsntation is still a valid liqu graph rprsntation of th distribution sin φ (V)φ (W) φ (S) = φ φ (V)φ (W) (S) φ(s) φ (S) = φ (V)φ (W) = p(x )

Absorption Aftr W absorbs information from V thn φ (W) ontains th marginal p(w). Similarly, aftr V absorbs information from W thn φ (V) ontains th marginal p(v). Aftr th sparator S has partiipatd in absorption along both dirtions, thn th sparator potntial will ontain p(s). Proof φ (S) = W\S φ (W) = W\S φ (W)φ (S) = Continuing, w hav th nw potntial φ (V) givn by φ (V) = φ (V)φ (S) φ (S) = {W V}\S = φ (V) W\S φ (W)φ (S)/ φ (S) W\S φ (V)φ (W) = p(v) φ (W)φ (V) = p(s)

Absorption Shdul on a Cliqu Tr A 3 4 B C 1 6 2 D 5 7 10 E 9 8 F For a valid shdul, mssags an only b passd to a nighbour whn all othr mssags hav bn rivd. Mor than on valid shdul may xist.

Forming a Cliqu Tr x 1 x 1, x 3 x 2 x 3, x 2, (a) (b) p(x 1, x 2, x 3, ) = φ(x 1, )φ(x 2, )φ(x 3, ) Th liqu graph of this singly-onntd Markov ntwork is multiply-onntd, whr th sparator potntials ar all st to unity. For absorption to work, w nd a singly-onntd liqu graph.

Forming a Cliqu Tr p(x 1, x 2, x 3, ) = φ(x 1, )φ(x 2, )φ(x 3, ) Rxprss th Markov ntwork in trms of marginals. First w hav th rlations p(x 1, ) = p(x 1, x 2, x 3, ) = φ(x 1, ) φ(x 2, ) φ(x 3, ) x 2,x 3 x 2 x 3 p(x 2, ) = p(x 1, x 2, x 3, ) = φ(x 2, ) φ(x 1, ) φ(x 3, ) x 1,x 3 x 1 x 3 p(x 3, ) = p(x 1, x 2, x 3, ) = φ(x 3, ) φ(x 1, ) φ(x 2, ) x 1,x 2 x 1 x 2 Taking th produt of th thr marginals, w hav p(x 1, )p(x 2, )p(x 3, ) ( ) 2 = φ(x 1, )φ(x 2, )φ(x 3, ) φ(x 1, ) φ(x 2, ) φ(x 3, ) } x 1 x 2 x 3 {{ } p() 2

Forming a Cliqu Tr This mans that th Markov ntwork an b xprssd in trms of marginals as p(x 1, x 2, x 3, ) = p(x 1, )p(x 2, )p(x 3, ) p( )p( ) Hn a valid liqu graph is also givn by x 1, x 1, x 3, x 2, x 3, x 2, If a variabl (hr ) ours on vry sparator in a liqu graph loop, on an rmov that variabl from an arbitrarily hosn sparator in th loop. Providd that th original Markov ntwork is singly-onntd, on an always form a liqu tr in this mannr.

Juntion Tr ab d f g h Running Intrstion Proprty A Cliqu Tr is a Juntion Tr if, for ah pair of nods, V and W, all nods on th path btwn V and W ontain th intrstion V W. Any singly-onntd Markov Ntwork an b transformd into a Juntion Tr. Thanks to th running intrstion proprty, loal onsistny of marginals propagats to global marginal onsistny.

Blif Nt Markov Nt a b a b d d f f g h g h Moralisation Form a link btwn all unmarrid parnts.

Markov Nt Juntion Tr ab ab d f d f g h g h Form th liqu graph Idntify a maximal wight spanning tr of th liqu graph. (Th wight of th dg is th numbr of variabls in th sparator)

Absorption a b ab d d f f g h g h Assign potntials to JT liqus. φ (ab) = p(a)p(b)p( a, b), φ (d) = p(d)p( d, ) φ (f) = p(f ), φ (g) = p(g ), φ (h) = p(h ) All sparator potntials ar initialisd to unity. Not that in som instans it an b that a juntion tr liqu is assignd to unity. Carry out absorption using a valid shdul. Marginals an thn b rad of th transformd potntials.