Stein s Method for concentration inequalities

Size: px
Start display at page:

Download "Stein s Method for concentration inequalities"

Transcription

1 Number of Triangles in Erdős-Rényi random graph, UC Berkeley joint work with Sourav Chatterjee, UC Berkeley Cornell Probability Summer School July 7, 2009

2 Number of triangles in Erdős-Rényi random graph Let T n be the number of triangles in G(n, p). Behavior of the upper tail of subgraph counts in G(n, p) is a problem of great interest in the theory of random graphs. Best known bounds due to Kim and Vu (2004) for triangles and Janson, Oleszkiewicz, and Ruciński (2004) for general subgraphs. For triangles the result of Kim and Vu essentially state that for any fixed ε > 0, e Θ(n2 p 2 log 1/p) P(T n (1 + ε) E[T n ]) e Θ(n2 p 2). We are interested in finding the function f (p, ε) such that P(T n (1 + ε) E[T n ]) = e n2 f (p,ε) (1+o(1)).

3 Large deviation result for number of triangles For r, p (0, 1), define I (r, p) := r log r p + (1 r) log 1 r 1 p. Theorem (Chatterjee & D.) Let T n be the number of triangles in G(n, p), where p > p 0 is fixed with p 0 = 2/(2 + e 3/2 ) Then for any r > p, ( P T n ( ) ] n )r 3 = exp [ n2 3 2 I (r, p) (1 + O(n 1/2 )). Moreover, if p p 0, there exist p, p (p, 1) such that the same result holds for all r (p, p ) (p, 1].

4 Figure: The set of (p, r) where our large deviation result holds.

5 Stein s method for concentration inequalities Let (X, X ) be an exchangeable pair of random variables. (X 0, X 1 ) is an exchangeable pair for a stationary reversible Markov chain X 0, X 1, X 2,... Let F (, ) be an antisymmetric function F (x, y) = F (y, x), such that E[F (X, X ) X ] = f (X ) with E[f (X )] = 0. By exchangeability and antisymmetry, for all g, we have E[g(X )f (X )] = E[g(X )F (X, X )] = 1 2 E[g(X )F (X, X ) + g(x )F (X, X )] = 1 2 E[(g(X ) g(x ))F (X, X )].

6 Concentration inequality using exchangeable pair In particular we have Var(f (X )) = 1 2 E[(f (X ) f (X ))F (X, X )]. Define (X ) := 1 2 E( f (X ) f (X ) F (X, X ) X ). Theorem (Chatterjee 05) Suppose (X ) Bf (X ) + C a.s., then for any t 0, ( P( f (X ) t) 2 exp t 2 2C + 2Bt ).

7 Concentration inequality using exchangeable pair In particular we have Var(f (X )) = 1 2 E[(f (X ) f (X ))F (X, X )]. Define (X ) := 1 2 E( f (X ) f (X ) F (X, X ) X ). Theorem (Chatterjee & D.) Suppose (X ) B f (X ) α + C a.s. for some real number α [0, 2), then for any t 0, ( P( f (X ) > t) 2 exp t 2 α 16 max{b, C 1 α/2 } ).

8 Towards the large deviation result... Any undirected graph on n vertices can be represented by x {0, 1} (n 2) where x ij = 1{edge (i, j) is present} for i < j and this correspondence is a bijection. We need to consider a Gibbs measure on the space of graphs on n vertices with Hamiltonian where T (x) = H(x) = β T (x) n 2 + he(x) i<j<kx ij x jk x ik and E(x) = i<j x ij. denote the number of triangles and edges in x respectively.

9 A new model? Let ψ n (β, h) := log x eh(x) denote log of the partition function. Note that for any β 0 we have ( ) n log P 0,h (T )r 3 β ( ) n r 3 + log E 0,h [e β 3 n 2 3 ( ) n r 3 = ψ n(β, h) ψ n (0, h). If β = 0 we get back G(n, p) with p = ψ n (0, h) = For β 0, ψ n (β, h) =? eh and 1+e h ( ) n log(1 + e h ). 2 n 2 T ]

10 Solution in the high temperature regime Define ϕ : [0, 1] [0, 1] by ϕ(x) = eβx+h 1 + e βx+h. Theorem (Chatterjee and D.) If (β, h) is in a high temperature regime, there is a unique p = p (β, h) [0, 1] such that p = ϕ(p 2 ) and ψ n (β, h) lim ( n n = 2) βp3 3 + hp p log p (1 p ) log(1 p ).

11 Main Step Define L ij := 1 n 2 the fraction of wedges at (i, j). X ik X jk k i,j Easy fact: E[ X ij rest of the edge configuration] = ϕ(l ij ). First step: For all β 0, h R we have ( ) L ij β ϕ(l ik )ϕ(l jk ) + O n 2 n with high probability. k i,j Proof uses suitable exchangeable pair using Glauber dynamics, suitable antisymmetric function and Stein s method for concentration inequalities theorem.

12 Other temperature free results Using similar techniques we have E(X) ( n ) ( 1 n ) ( ) 1 + β ϕ(l ij ) + O n 2 2 i<j T (X) ( n ) ( 1 n ) ( ) 1 + β ϕ(l ij )ϕ(l jk )ϕ(l ik ) + O. n 3 3 i<j<k

13 Rigorous result in high temperature regime When (β, h) is in high temperature regime, the system of equations a ij = 1 ϕ(a ik )ϕ(a jk ) n 2 k i,j has a unique solution at (a ij = u for all i < j). Lemma (High temperature regime) Let β, h be such that the equation u = ϕ(u) 2 has a unique solution at u. Assume that 2ϕ(u )ϕ (u ) < 1. Then for each i < j, we have E L ij u K(β, h) n where K(β, h) is a constant depending only on β, h.

14 High temperature regime Figure: The shaded region corresponds to the high temperature regime. This boundary was predicted to be the phase transition curve using physical arguments by Park and Newman (2005).

15 Partition function in high temperature regime Temperature free results then imply that ( ) E n E(X) p K(β, h)n 3/2 2 ( ) E n T (X) p 3 3 K(β, h)n5/2 where p = ϕ(u ). We have H(x) = β T (x) n 2 +he(x) where h is such that p = Hence ψ n (β, h) ( ) ( ) n βp (h h )p +h E(x) eh 1+e. h ( ) ( ) n βp (h h )p + ψ n (0, h ).

16 Proof of the Large deviation result So given p, r find β such that r = p (β, h) where p = The upper bound holds. The lower bound can be proved by showing that ( ) n P( T n )r 3 C(p, r)n 5/2 3 ( ) = exp n2 I (r, p) (1 + O(n 1/2 )) 2 for r, p as above. eh. 1+e h

17 More Applications Similar results for general subgraph counts in G(n, p). Optimal tail bounds for magnetization in Curie-Weiss model at critical temperature. Tail bounds for magnetization in Ising model on Z d.

18 Open Questions What happens in the low temperature regime for the exponential random graph model? Partial results are available. Extend the large deviation result to the rest of the (p, r) region. A complete solution for the exponential random graph is not enough.

19 Thank you!

Concentration inequalities for non-lipschitz functions

Concentration inequalities for non-lipschitz functions Concentration inequalities for non-lipschitz functions University of Warsaw Berkeley, October 1, 2013 joint work with Radosław Adamczak (University of Warsaw) Gaussian concentration (Sudakov-Tsirelson,

More information

arxiv: v1 [math.pr] 11 Dec 2008

arxiv: v1 [math.pr] 11 Dec 2008 Mixing time of exponential random graphs Shankar Bhamidi Guy Bresler Allan Sly arxiv:081.65v1 [math.pr] 11 Dec 008 Abstract A variety of random graph models have been developed in recent years to study

More information

The large deviation principle for the Erdős-Rényi random graph

The large deviation principle for the Erdős-Rényi random graph The large deviation principle for the Erdős-Rényi random graph (Courant Institute, NYU) joint work with S. R. S. Varadhan Main objective: how to count graphs with a given property Only consider finite

More information

Subhypergraph counts in extremal and random hypergraphs and the fractional q-independence

Subhypergraph counts in extremal and random hypergraphs and the fractional q-independence Subhypergraph counts in extremal and random hypergraphs and the fractional q-independence Andrzej Dudek adudek@emory.edu Andrzej Ruciński rucinski@amu.edu.pl June 21, 2008 Joanna Polcyn joaska@amu.edu.pl

More information

Concentration of Measures by Bounded Size Bias Couplings

Concentration of Measures by Bounded Size Bias Couplings Concentration of Measures by Bounded Size Bias Couplings Subhankar Ghosh, Larry Goldstein University of Southern California [arxiv:0906.3886] January 10 th, 2013 Concentration of Measure Distributional

More information

Stein Couplings for Concentration of Measure

Stein Couplings for Concentration of Measure Stein Couplings for Concentration of Measure Jay Bartroff, Subhankar Ghosh, Larry Goldstein and Ümit Işlak University of Southern California [arxiv:0906.3886] [arxiv:1304.5001] [arxiv:1402.6769] Borchard

More information

Concentration of Measures by Bounded Couplings

Concentration of Measures by Bounded Couplings Concentration of Measures by Bounded Couplings Subhankar Ghosh, Larry Goldstein and Ümit Işlak University of Southern California [arxiv:0906.3886] [arxiv:1304.5001] May 2013 Concentration of Measure Distributional

More information

7.1 Coupling from the Past

7.1 Coupling from the Past Georgia Tech Fall 2006 Markov Chain Monte Carlo Methods Lecture 7: September 12, 2006 Coupling from the Past Eric Vigoda 7.1 Coupling from the Past 7.1.1 Introduction We saw in the last lecture how Markov

More information

Spin glasses and Stein s method

Spin glasses and Stein s method UC Berkeley Spin glasses Magnetic materials with strange behavior. ot explained by ferromagnetic models like the Ising model. Theoretically studied since the 70 s. An important example: Sherrington-Kirkpatrick

More information

Lecture 7: The Subgraph Isomorphism Problem

Lecture 7: The Subgraph Isomorphism Problem CSC2429, MAT1304: Circuit Complexity October 25, 2016 Lecture 7: The Subgraph Isomorphism Problem Instructor: Benjamin Rossman 1 The Problem SUB(G) Convention 1 (Graphs). Graphs are finite simple graphs

More information

Glauber Dynamics for Ising Model I AMS Short Course

Glauber Dynamics for Ising Model I AMS Short Course Glauber Dynamics for Ising Model I AMS Short Course January 2010 Ising model Let G n = (V n, E n ) be a graph with N = V n < vertices. The nearest-neighbor Ising model on G n is the probability distribution

More information

Ferromagnets and the classical Heisenberg model. Kay Kirkpatrick, UIUC

Ferromagnets and the classical Heisenberg model. Kay Kirkpatrick, UIUC Ferromagnets and the classical Heisenberg model Kay Kirkpatrick, UIUC Ferromagnets and the classical Heisenberg model: asymptotics for a mean-field phase transition Kay Kirkpatrick, Urbana-Champaign June

More information

Small subgraphs of random regular graphs

Small subgraphs of random regular graphs Discrete Mathematics 307 (2007 1961 1967 Note Small subgraphs of random regular graphs Jeong Han Kim a,b, Benny Sudakov c,1,vanvu d,2 a Theory Group, Microsoft Research, Redmond, WA 98052, USA b Department

More information

arxiv: v2 [math.pr] 22 Oct 2018

arxiv: v2 [math.pr] 22 Oct 2018 Breaking of ensemble equivalence for perturbed Erdős-Rényi random graphs arxiv:807.07750v [math.pr] Oct 08 F. den Hollander M. Mandjes A. Roccaverde 3 N.J. Starreveld 4 October 3, 08 Abstract In a previous

More information

Mod-φ convergence II: dependency graphs

Mod-φ convergence II: dependency graphs Mod-φ convergence II: dependency graphs Valentin Féray (joint work with Pierre-Loïc Méliot and Ashkan Nikeghbali) Institut für Mathematik, Universität Zürich Summer school in Villa Volpi, Lago Maggiore,

More information

The typical structure of sparse K r+1 -free graphs

The typical structure of sparse K r+1 -free graphs The typical structure of sparse K r+1 -free graphs Lutz Warnke University of Cambridge (joint work with József Balogh, Robert Morris, and Wojciech Samotij) H-free graphs / Turán s theorem Definition Let

More information

Graph Limits: Some Open Problems

Graph Limits: Some Open Problems Graph Limits: Some Open Problems 1 Introduction Here are some questions from the open problems session that was held during the AIM Workshop Graph and Hypergraph Limits, Palo Alto, August 15-19, 2011.

More information

Notes 6 : First and second moment methods

Notes 6 : First and second moment methods Notes 6 : First and second moment methods Math 733-734: Theory of Probability Lecturer: Sebastien Roch References: [Roc, Sections 2.1-2.3]. Recall: THM 6.1 (Markov s inequality) Let X be a non-negative

More information

Ground states for exponential random graphs

Ground states for exponential random graphs Ground states for exponential random graphs Mei Yin Department of Mathematics, University of Denver March 5, 2018 We propose a perturbative method to estimate the normalization constant in exponential

More information

A note on perfect simulation for exponential random graph models

A note on perfect simulation for exponential random graph models A note on perfect simulation for exponential random graph models A. Cerqueira, A. Garivier and F. Leonardi October 4, 017 arxiv:1710.00873v1 [stat.co] Oct 017 Abstract In this paper we propose a perfect

More information

arxiv: v3 [math.pr] 19 Apr 2018

arxiv: v3 [math.pr] 19 Apr 2018 Exponential random graphs behave like mixtures of stochastic block models Ronen Eldan and Renan Gross arxiv:70707v3 [mathpr] 9 Apr 08 Abstract We study the behavior of exponential random graphs in both

More information

Rapid Introduction to Machine Learning/ Deep Learning

Rapid Introduction to Machine Learning/ Deep Learning Rapid Introduction to Machine Learning/ Deep Learning Hyeong In Choi Seoul National University 1/24 Lecture 5b Markov random field (MRF) November 13, 2015 2/24 Table of contents 1 1. Objectives of Lecture

More information

Solutions to Problem Set 5

Solutions to Problem Set 5 UC Berkeley, CS 74: Combinatorics and Discrete Probability (Fall 00 Solutions to Problem Set (MU 60 A family of subsets F of {,,, n} is called an antichain if there is no pair of sets A and B in F satisfying

More information

Wasserstein-2 bounds in normal approximation under local dependence

Wasserstein-2 bounds in normal approximation under local dependence Wasserstein- bounds in normal approximation under local dependence arxiv:1807.05741v1 [math.pr] 16 Jul 018 Xiao Fang The Chinese University of Hong Kong Abstract: We obtain a general bound for the Wasserstein-

More information

Stein s Method: Distributional Approximation and Concentration of Measure

Stein s Method: Distributional Approximation and Concentration of Measure Stein s Method: Distributional Approximation and Concentration of Measure Larry Goldstein University of Southern California 36 th Midwest Probability Colloquium, 2014 Concentration of Measure Distributional

More information

Lecture 3: Expected Value. These integrals are taken over all of Ω. If we wish to integrate over a measurable subset A Ω, we will write

Lecture 3: Expected Value. These integrals are taken over all of Ω. If we wish to integrate over a measurable subset A Ω, we will write Lecture 3: Expected Value 1.) Definitions. If X 0 is a random variable on (Ω, F, P), then we define its expected value to be EX = XdP. Notice that this quantity may be. For general X, we say that EX exists

More information

Contributions to Stein s method and some limit theorems in probability. Partha Sarathi Dey

Contributions to Stein s method and some limit theorems in probability. Partha Sarathi Dey Contributions to Stein s method and some limit theorems in probability by Partha Sarathi Dey A dissertation submitted in partial satisfaction of the requirements for the degree of Doctor of Philosophy

More information

CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash

CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash Equilibrium Price of Stability Coping With NP-Hardness

More information

Mixing in Product Spaces. Elchanan Mossel

Mixing in Product Spaces. Elchanan Mossel Poincaré Recurrence Theorem Theorem (Poincaré, 1890) Let f : X X be a measure preserving transformation. Let E X measurable. Then P[x E : f n (x) / E, n > N(x)] = 0 Poincaré Recurrence Theorem Theorem

More information

Vertex colorings of graphs without short odd cycles

Vertex colorings of graphs without short odd cycles Vertex colorings of graphs without short odd cycles Andrzej Dudek and Reshma Ramadurai Department of Mathematical Sciences Carnegie Mellon University Pittsburgh, PA 1513, USA {adudek,rramadur}@andrew.cmu.edu

More information

arxiv: v3 [math.pr] 6 Apr 2011

arxiv: v3 [math.pr] 6 Apr 2011 ESTIMATING AND UNDERSTANDING EXPONENTIAL RANDOM GRAPH MODELS SOURAV CHATTERJEE AND PERSI DIACONIS arxiv:1102.2650v3 [math.pr] 6 Apr 2011 Abstract. We introduce a new method for estimating the parameters

More information

RANDOM GRAPHS BY JOEL SPENCER. as n.

RANDOM GRAPHS BY JOEL SPENCER. as n. RANDOM GRAHS BY JOEL SENCER Notation 1. We say fn gn as n if fx gx as n. Notation 2. We say fn gn as n if fn gn 1 as n Notation 3. We say fn = ogn as n if fn gn 0 as n Notation 4. n k := nn 1n 2... n k

More information

ESTIMATING AND UNDERSTANDING EXPONENTIAL RANDOM GRAPH MODELS

ESTIMATING AND UNDERSTANDING EXPONENTIAL RANDOM GRAPH MODELS ESTIMATING AND UNDERSTANDING EXPONENTIAL RANDOM GRAPH MODELS SOURAV CHATTERJEE AND PERSI DIACONIS Abstract. We introduce a new method for estimating the parameters of exponential random graph models. The

More information

Ferromagnets and superconductors. Kay Kirkpatrick, UIUC

Ferromagnets and superconductors. Kay Kirkpatrick, UIUC Ferromagnets and superconductors Kay Kirkpatrick, UIUC Ferromagnet and superconductor models: Phase transitions and asymptotics Kay Kirkpatrick, Urbana-Champaign October 2012 Ferromagnet and superconductor

More information

Statistical and Computational Phase Transitions in Planted Models

Statistical and Computational Phase Transitions in Planted Models Statistical and Computational Phase Transitions in Planted Models Jiaming Xu Joint work with Yudong Chen (UC Berkeley) Acknowledgement: Prof. Bruce Hajek November 4, 203 Cluster/Community structure in

More information

Approximate counting of large subgraphs in random graphs with statistical mechanics methods

Approximate counting of large subgraphs in random graphs with statistical mechanics methods Approximate counting of large subgraphs in random graphs with statistical mechanics methods Guilhem Semerjian LPT-ENS Paris 13.03.08 / Eindhoven in collaboration with Rémi Monasson, Enzo Marinari and Valery

More information

A LARGE DEVIATION PRINCIPLE FOR THE ERDŐS RÉNYI UNIFORM RANDOM GRAPH

A LARGE DEVIATION PRINCIPLE FOR THE ERDŐS RÉNYI UNIFORM RANDOM GRAPH A LARGE DEVIATION PRINCIPLE FOR THE ERDŐS RÉNYI UNIFORM RANDOM GRAPH AMIR DEMBO AND EYAL LUBETZKY Abstract. Starting with the large deviation principle (ldp) for the Erdős Rényi binomial random graph G(n,

More information

FIRST ORDER SENTENCES ON G(n, p), ZERO-ONE LAWS, ALMOST SURE AND COMPLETE THEORIES ON SPARSE RANDOM GRAPHS

FIRST ORDER SENTENCES ON G(n, p), ZERO-ONE LAWS, ALMOST SURE AND COMPLETE THEORIES ON SPARSE RANDOM GRAPHS FIRST ORDER SENTENCES ON G(n, p), ZERO-ONE LAWS, ALMOST SURE AND COMPLETE THEORIES ON SPARSE RANDOM GRAPHS MOUMANTI PODDER 1. First order theory on G(n, p) We start with a very simple property of G(n,

More information

2 The Curie Weiss Model

2 The Curie Weiss Model 2 The Curie Weiss Model In statistical mechanics, a mean-field approximation is often used to approximate a model by a simpler one, whose global behavior can be studied with the help of explicit computations.

More information

Asymptotic Behavior of the Magnetization Near Critical and Tricritical Points via Ginzburg-Landau Polynomials

Asymptotic Behavior of the Magnetization Near Critical and Tricritical Points via Ginzburg-Landau Polynomials Asymptotic Behavior of the Magnetization Near Critical and Tricritical Points via Ginzburg-Landau Polynomials Richard S. Ellis rsellis@math.umass.edu Jonathan Machta 2 machta@physics.umass.edu Peter Tak-Hun

More information

BIRTHDAY PROBLEM, MONOCHROMATIC SUBGRAPHS & THE SECOND MOMENT PHENOMENON / 23

BIRTHDAY PROBLEM, MONOCHROMATIC SUBGRAPHS & THE SECOND MOMENT PHENOMENON / 23 BIRTHDAY PROBLEM, MONOCHROMATIC SUBGRAPHS & THE SECOND MOMENT PHENOMENON Somabha Mukherjee 1 University of Pennsylvania March 30, 2018 Joint work with Bhaswar B. Bhattacharya 2 and Sumit Mukherjee 3 1

More information

CONCENTRATION OF MULTIVARIATE POLYNOMIALS AND ITS APPLICATIONS. Jeong Han Kim Microsoft Research One Microsoft Way Redmond WA Van H.

CONCENTRATION OF MULTIVARIATE POLYNOMIALS AND ITS APPLICATIONS. Jeong Han Kim Microsoft Research One Microsoft Way Redmond WA Van H. CONCENTRATION OF MULTIVARIATE POLYNOMIALS AND ITS APPLICATIONS Jeong Han Kim Microsoft Research One Microsoft Way Redmond WA98052 Van H. Vu Microsoft Research One Microsoft Way Redmond WA98052 Abstract.

More information

Exponential families also behave nicely under conditioning. Specifically, suppose we write η = (η 1, η 2 ) R k R p k so that

Exponential families also behave nicely under conditioning. Specifically, suppose we write η = (η 1, η 2 ) R k R p k so that 1 More examples 1.1 Exponential families under conditioning Exponential families also behave nicely under conditioning. Specifically, suppose we write η = η 1, η 2 R k R p k so that dp η dm 0 = e ηt 1

More information

Generalized Exponential Random Graph Models: Inference for Weighted Graphs

Generalized Exponential Random Graph Models: Inference for Weighted Graphs Generalized Exponential Random Graph Models: Inference for Weighted Graphs James D. Wilson University of North Carolina at Chapel Hill June 18th, 2015 Political Networks, 2015 James D. Wilson GERGMs for

More information

Disjointness and Additivity

Disjointness and Additivity Midterm 2: Format Midterm 2 Review CS70 Summer 2016 - Lecture 6D David Dinh 28 July 2016 UC Berkeley 8 questions, 190 points, 110 minutes (same as MT1). Two pages (one double-sided sheet) of handwritten

More information

Mod-φ convergence I: examples and probabilistic estimates

Mod-φ convergence I: examples and probabilistic estimates Mod-φ convergence I: examples and probabilistic estimates Valentin Féray (joint work with Pierre-Loïc Méliot and Ashkan Nikeghbali) Institut für Mathematik, Universität Zürich Summer school in Villa Volpi,

More information

Markov Random Fields

Markov Random Fields Markov Random Fields 1. Markov property The Markov property of a stochastic sequence {X n } n 0 implies that for all n 1, X n is independent of (X k : k / {n 1, n, n + 1}), given (X n 1, X n+1 ). Another

More information

Szemerédi s regularity lemma revisited. Lewis Memorial Lecture March 14, Terence Tao (UCLA)

Szemerédi s regularity lemma revisited. Lewis Memorial Lecture March 14, Terence Tao (UCLA) Szemerédi s regularity lemma revisited Lewis Memorial Lecture March 14, 2008 Terence Tao (UCLA) 1 Finding models of large dense graphs Suppose we are given a large dense graph G = (V, E), where V is a

More information

Metastability for the Ising model on a random graph

Metastability for the Ising model on a random graph Metastability for the Ising model on a random graph Joint project with Sander Dommers, Frank den Hollander, Francesca Nardi Outline A little about metastability Some key questions A little about the, and

More information

RANDOM GRAPHS. Joel Spencer ICTP 12

RANDOM GRAPHS. Joel Spencer ICTP 12 RANDOM GRAPHS Joel Spencer ICTP 12 Graph Theory Preliminaries A graph G, formally speaking, is a pair (V (G), E(G)) where the elements v V (G) are called vertices and the elements of E(G), called edges,

More information

Other Things & Some Applications

Other Things & Some Applications Chapter 4 Other Things & Some Applications 4. Unimodality, Log-concavity & Real-rootedness (4.) Denition Let A = (a 0, a,..., a n ) be a sequence of positive real numbers. The sequence A is said to be:

More information

Kolmogorov Berry-Esseen bounds for binomial functionals

Kolmogorov Berry-Esseen bounds for binomial functionals Kolmogorov Berry-Esseen bounds for binomial functionals Raphaël Lachièze-Rey, Univ. South California, Univ. Paris 5 René Descartes, Joint work with Giovanni Peccati, University of Luxembourg, Singapore

More information

Midterm 2 Review. CS70 Summer Lecture 6D. David Dinh 28 July UC Berkeley

Midterm 2 Review. CS70 Summer Lecture 6D. David Dinh 28 July UC Berkeley Midterm 2 Review CS70 Summer 2016 - Lecture 6D David Dinh 28 July 2016 UC Berkeley Midterm 2: Format 8 questions, 190 points, 110 minutes (same as MT1). Two pages (one double-sided sheet) of handwritten

More information

CHAPTER 4. Cluster expansions

CHAPTER 4. Cluster expansions CHAPTER 4 Cluster expansions The method of cluster expansions allows to write the grand-canonical thermodynamic potential as a convergent perturbation series, where the small parameter is related to the

More information

Computing bounds for entropy of stationary Z d -Markov random fields. Brian Marcus (University of British Columbia)

Computing bounds for entropy of stationary Z d -Markov random fields. Brian Marcus (University of British Columbia) Computing bounds for entropy of stationary Z d -Markov random fields Brian Marcus (University of British Columbia) Ronnie Pavlov (University of Denver) WCI 2013 Hong Kong University Dec. 11, 2013 1 Outline

More information

Susceptible-Infective-Removed Epidemics and Erdős-Rényi random

Susceptible-Infective-Removed Epidemics and Erdős-Rényi random Susceptible-Infective-Removed Epidemics and Erdős-Rényi random graphs MSR-Inria Joint Centre October 13, 2015 SIR epidemics: the Reed-Frost model Individuals i [n] when infected, attempt to infect all

More information

Structured Variational Inference

Structured Variational Inference Structured Variational Inference Sargur srihari@cedar.buffalo.edu 1 Topics 1. Structured Variational Approximations 1. The Mean Field Approximation 1. The Mean Field Energy 2. Maximizing the energy functional:

More information

Antiferromagnetic Potts models and random colorings

Antiferromagnetic Potts models and random colorings Antiferromagnetic Potts models and random colorings of planar graphs. joint with A.D. Sokal (New York) and R. Kotecký (Prague) October 9, 0 Gibbs measures Let G = (V, E) be a finite graph and let S be

More information

Int. Statistical Inst.: Proc. 58th World Statistical Congress, 2011, Dublin (Session CPS014) p.4149

Int. Statistical Inst.: Proc. 58th World Statistical Congress, 2011, Dublin (Session CPS014) p.4149 Int. Statistical Inst.: Proc. 58th orld Statistical Congress, 011, Dublin (Session CPS014) p.4149 Invariant heory for Hypothesis esting on Graphs Priebe, Carey Johns Hopkins University, Applied Mathematics

More information

The expansion of random regular graphs

The expansion of random regular graphs The expansion of random regular graphs David Ellis Introduction Our aim is now to show that for any d 3, almost all d-regular graphs on {1, 2,..., n} have edge-expansion ratio at least c d d (if nd is

More information

LECTURE 3. Last time:

LECTURE 3. Last time: LECTURE 3 Last time: Mutual Information. Convexity and concavity Jensen s inequality Information Inequality Data processing theorem Fano s Inequality Lecture outline Stochastic processes, Entropy rate

More information

3 Undirected Graphical Models

3 Undirected Graphical Models Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science 6.438 Algorithms For Inference Fall 2014 3 Undirected Graphical Models In this lecture, we discuss undirected

More information

Phase transitions in the edge-triangle exponential random graph model

Phase transitions in the edge-triangle exponential random graph model Phase transitions in the edge-triangle exponential random graph model Mei Yin Department of Mathematics, Brown University April 28, 2014 Probability space: The set G n of all simple graphs G n on n vertices.

More information

Markov Chains and MCMC

Markov Chains and MCMC Markov Chains and MCMC Markov chains Let S = {1, 2,..., N} be a finite set consisting of N states. A Markov chain Y 0, Y 1, Y 2,... is a sequence of random variables, with Y t S for all points in time

More information

Networks: Lectures 9 & 10 Random graphs

Networks: Lectures 9 & 10 Random graphs Networks: Lectures 9 & 10 Random graphs Heather A Harrington Mathematical Institute University of Oxford HT 2017 What you re in for Week 1: Introduction and basic concepts Week 2: Small worlds Week 3:

More information

Edge-disjoint induced subgraphs with given minimum degree

Edge-disjoint induced subgraphs with given minimum degree Edge-disjoint induced subgraphs with given minimum degree Raphael Yuster Department of Mathematics University of Haifa Haifa 31905, Israel raphy@math.haifa.ac.il Submitted: Nov 9, 01; Accepted: Feb 5,

More information

Random Graphs. 7.1 Introduction

Random Graphs. 7.1 Introduction 7 Random Graphs 7.1 Introduction The theory of random graphs began in the late 1950s with the seminal paper by Erdös and Rényi [?]. In contrast to percolation theory, which emerged from efforts to model

More information

Minimal basis for connected Markov chain over 3 3 K contingency tables with fixed two-dimensional marginals. Satoshi AOKI and Akimichi TAKEMURA

Minimal basis for connected Markov chain over 3 3 K contingency tables with fixed two-dimensional marginals. Satoshi AOKI and Akimichi TAKEMURA Minimal basis for connected Markov chain over 3 3 K contingency tables with fixed two-dimensional marginals Satoshi AOKI and Akimichi TAKEMURA Graduate School of Information Science and Technology University

More information

Intelligent Systems:

Intelligent Systems: Intelligent Systems: Undirected Graphical models (Factor Graphs) (2 lectures) Carsten Rother 15/01/2015 Intelligent Systems: Probabilistic Inference in DGM and UGM Roadmap for next two lectures Definition

More information

The symmetry in the martingale inequality

The symmetry in the martingale inequality Statistics & Probability Letters 56 2002 83 9 The symmetry in the martingale inequality Sungchul Lee a; ;, Zhonggen Su a;b;2 a Department of Mathematics, Yonsei University, Seoul 20-749, South Korea b

More information

EMERGENT STRUCTURES IN LARGE NETWORKS

EMERGENT STRUCTURES IN LARGE NETWORKS Applied Probability Trust (29 November 2012) EMERGENT STRUCTURES IN LARGE NETWORKS DAVID ARISTOFF, University of Minnesota CHARLES RADIN, The University of Texas at Austin Abstract We consider a large

More information

The Ising Partition Function: Zeros and Deterministic Approximation

The Ising Partition Function: Zeros and Deterministic Approximation The Ising Partition Function: Zeros and Deterministic Approximation Jingcheng Liu Alistair Sinclair Piyush Srivastava University of California, Berkeley Summer 2017 Jingcheng Liu (UC Berkeley) The Ising

More information

STAT 200C: High-dimensional Statistics

STAT 200C: High-dimensional Statistics STAT 200C: High-dimensional Statistics Arash A. Amini May 30, 2018 1 / 59 Classical case: n d. Asymptotic assumption: d is fixed and n. Basic tools: LLN and CLT. High-dimensional setting: n d, e.g. n/d

More information

Undirected Graphical Models

Undirected Graphical Models Undirected Graphical Models 1 Conditional Independence Graphs Let G = (V, E) be an undirected graph with vertex set V and edge set E, and let A, B, and C be subsets of vertices. We say that C separates

More information

EE/Stats 376A: Homework 7 Solutions Due on Friday March 17, 5 pm

EE/Stats 376A: Homework 7 Solutions Due on Friday March 17, 5 pm EE/Stats 376A: Homework 7 Solutions Due on Friday March 17, 5 pm 1. Feedback does not increase the capacity. Consider a channel with feedback. We assume that all the recieved outputs are sent back immediately

More information

Quasi-randomness of graph balanced cut properties

Quasi-randomness of graph balanced cut properties Quasi-randomness of graph balanced cut properties Hao Huang Choongbum Lee Abstract Quasi-random graphs can be informally described as graphs whose edge distribution closely resembles that of a truly random

More information

L p Spaces and Convexity

L p Spaces and Convexity L p Spaces and Convexity These notes largely follow the treatments in Royden, Real Analysis, and Rudin, Real & Complex Analysis. 1. Convex functions Let I R be an interval. For I open, we say a function

More information

Lecture 7: February 6

Lecture 7: February 6 CS271 Randomness & Computation Spring 2018 Instructor: Alistair Sinclair Lecture 7: February 6 Disclaimer: These notes have not been subjected to the usual scrutiny accorded to formal publications. They

More information

MONOTONE COUPLING AND THE ISING MODEL

MONOTONE COUPLING AND THE ISING MODEL MONOTONE COUPLING AND THE ISING MODEL 1. PERFECT MATCHING IN BIPARTITE GRAPHS Definition 1. A bipartite graph is a graph G = (V, E) whose vertex set V can be partitioned into two disjoint set V I, V O

More information

Evolutionary Dynamics on Graphs

Evolutionary Dynamics on Graphs Evolutionary Dynamics on Graphs Leslie Ann Goldberg, University of Oxford Absorption Time of the Moran Process (2014) with Josep Díaz, David Richerby and Maria Serna Approximating Fixation Probabilities

More information

A CLASSROOM NOTE: ENTROPY, INFORMATION, AND MARKOV PROPERTY. Zoran R. Pop-Stojanović. 1. Introduction

A CLASSROOM NOTE: ENTROPY, INFORMATION, AND MARKOV PROPERTY. Zoran R. Pop-Stojanović. 1. Introduction THE TEACHING OF MATHEMATICS 2006, Vol IX,, pp 2 A CLASSROOM NOTE: ENTROPY, INFORMATION, AND MARKOV PROPERTY Zoran R Pop-Stojanović Abstract How to introduce the concept of the Markov Property in an elementary

More information

Reverse Hyper Contractive Inequalities: What? Why?? When??? How????

Reverse Hyper Contractive Inequalities: What? Why?? When??? How???? Reverse Hyper Contractive Inequalities: What? Why?? When??? How???? UC Berkeley Warsaw Cambridge May 21, 2012 Hyper-Contractive Inequalities The noise (Bonami-Beckner) semi-group on { 1, 1} n 1 2 Def 1:

More information

CONDITIONED POISSON DISTRIBUTIONS AND THE CONCENTRATION OF CHROMATIC NUMBERS

CONDITIONED POISSON DISTRIBUTIONS AND THE CONCENTRATION OF CHROMATIC NUMBERS CONDITIONED POISSON DISTRIBUTIONS AND THE CONCENTRATION OF CHROMATIC NUMBERS JOHN HARTIGAN, DAVID POLLARD AND SEKHAR TATIKONDA YALE UNIVERSITY ABSTRACT. The paper provides a simpler method for proving

More information

CPSC 540: Machine Learning

CPSC 540: Machine Learning CPSC 540: Machine Learning More Approximate Inference Mark Schmidt University of British Columbia Winter 2018 Last Time: Approximate Inference We ve been discussing graphical models for density estimation,

More information

Ergodic Theorems. Samy Tindel. Purdue University. Probability Theory 2 - MA 539. Taken from Probability: Theory and examples by R.

Ergodic Theorems. Samy Tindel. Purdue University. Probability Theory 2 - MA 539. Taken from Probability: Theory and examples by R. Ergodic Theorems Samy Tindel Purdue University Probability Theory 2 - MA 539 Taken from Probability: Theory and examples by R. Durrett Samy T. Ergodic theorems Probability Theory 1 / 92 Outline 1 Definitions

More information

The Transition Probability Function P ij (t)

The Transition Probability Function P ij (t) The Transition Probability Function P ij (t) Consider a continuous time Markov chain {X(t), t 0}. We are interested in the probability that in t time units the process will be in state j, given that it

More information

Random regular digraphs: singularity and spectrum

Random regular digraphs: singularity and spectrum Random regular digraphs: singularity and spectrum Nick Cook, UCLA Probability Seminar, Stanford University November 2, 2015 Universality Circular law Singularity probability Talk outline 1 Universality

More information

10708 Graphical Models: Homework 2

10708 Graphical Models: Homework 2 10708 Graphical Models: Homework 2 Due Monday, March 18, beginning of class Feburary 27, 2013 Instructions: There are five questions (one for extra credit) on this assignment. There is a problem involves

More information

SPIN SYSTEMS: HARDNESS OF APPROXIMATE COUNTING VIA PHASE TRANSITIONS

SPIN SYSTEMS: HARDNESS OF APPROXIMATE COUNTING VIA PHASE TRANSITIONS SPIN SYSTEMS: HARDNESS OF APPROXIMATE COUNTING VIA PHASE TRANSITIONS Andreas Galanis Joint work with: Jin-Yi Cai Leslie Ann Goldberg Heng Guo Mark Jerrum Daniel Štefankovič Eric Vigoda The Power of Randomness

More information

Hard-Core Model on Random Graphs

Hard-Core Model on Random Graphs Hard-Core Model on Random Graphs Antar Bandyopadhyay Theoretical Statistics and Mathematics Unit Seminar Theoretical Statistics and Mathematics Unit Indian Statistical Institute, New Delhi Centre New Delhi,

More information

arxiv: v1 [math.pr] 3 Apr 2017

arxiv: v1 [math.pr] 3 Apr 2017 KOLMOGOROV BOUNDS FOR THE NORMAL APPROXIMATION OF THE NUMBER OF TRIANGLES IN THE ERDŐS-RÉNYI RANDOM GRAPH Adrian Röllin National University of Singapore arxiv:70.000v [math.pr] Apr 07 Abstract We bound

More information

ON THE NUMBER OF ALTERNATING PATHS IN BIPARTITE COMPLETE GRAPHS

ON THE NUMBER OF ALTERNATING PATHS IN BIPARTITE COMPLETE GRAPHS ON THE NUMBER OF ALTERNATING PATHS IN BIPARTITE COMPLETE GRAPHS PATRICK BENNETT, ANDRZEJ DUDEK, ELLIOT LAFORGE December 1, 016 Abstract. Let C [r] m be a code such that any two words of C have Hamming

More information

Information Theory. Lecture 5 Entropy rate and Markov sources STEFAN HÖST

Information Theory. Lecture 5 Entropy rate and Markov sources STEFAN HÖST Information Theory Lecture 5 Entropy rate and Markov sources STEFAN HÖST Universal Source Coding Huffman coding is optimal, what is the problem? In the previous coding schemes (Huffman and Shannon-Fano)it

More information

Zig-Zag Monte Carlo. Delft University of Technology. Joris Bierkens February 7, 2017

Zig-Zag Monte Carlo. Delft University of Technology. Joris Bierkens February 7, 2017 Zig-Zag Monte Carlo Delft University of Technology Joris Bierkens February 7, 2017 Joris Bierkens (TU Delft) Zig-Zag Monte Carlo February 7, 2017 1 / 33 Acknowledgements Collaborators Andrew Duncan Paul

More information

De Finetti theorems for a Boolean analogue of easy quantum groups

De Finetti theorems for a Boolean analogue of easy quantum groups De Finetti theorems for a Boolean analogue of easy quantum groups Tomohiro Hayase Graduate School of Mathematical Sciences, the University of Tokyo March, 2016 Free Probability and the Large N limit, V

More information

ACO Comprehensive Exam October 14 and 15, 2013

ACO Comprehensive Exam October 14 and 15, 2013 1. Computability, Complexity and Algorithms (a) Let G be the complete graph on n vertices, and let c : V (G) V (G) [0, ) be a symmetric cost function. Consider the following closest point heuristic for

More information

linear programming and approximate constraint satisfaction

linear programming and approximate constraint satisfaction linear programming and approximate constraint satisfaction Siu On Chan MSR New England James R. Lee University of Washington Prasad Raghavendra UC Berkeley David Steurer Cornell results MAIN THEOREM: Any

More information

Gärtner-Ellis Theorem and applications.

Gärtner-Ellis Theorem and applications. Gärtner-Ellis Theorem and applications. Elena Kosygina July 25, 208 In this lecture we turn to the non-i.i.d. case and discuss Gärtner-Ellis theorem. As an application, we study Curie-Weiss model with

More information

A zero-one law for the existence of triangles in random key graphs

A zero-one law for the existence of triangles in random key graphs THE INSTITUTE FOR SYSTEMS RESEARCH ISR TECHNICAL REPORT 2011-11 A zero-one law for the existence of triangles in random key graphs Osman Yagan and Armand M. Makowski ISR develops, applies and teaches advanced

More information

An Application of First-Order Logic to a Problem in Combinatorics 1

An Application of First-Order Logic to a Problem in Combinatorics 1 An Application of First-Order Logic to a Problem in Combinatorics 1 I. The Combinatorial Background. A. Families of objects. In combinatorics, one frequently encounters a set of objects in which a), each

More information