A stretched-exponential Pinning Model with heavy-tailed Disorder

Size: px
Start display at page:

Download "A stretched-exponential Pinning Model with heavy-tailed Disorder"

Transcription

1 A stretched-exponential Pinning Model with heavy-tailed Disorder Niccolò Torri Université Claude-Bernard - Lyon 1 & Università degli Studi di Milano-Bicocca Roma, October 8, 2014

2 References A. Auffinger and O. Louidor Directed polymers in random environment with heavy tails Comm. on Pure and Applied Math., 64: , 2011 B. Hambly and J. B. Martin Heavy tails in last-passage percolation Probability Theory and Related Fields, 137: , 2007 N. Torri Pinning Model with Heavy Tailed disorder arxiv: , 2014 N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

3 1 Overview 2 Definitions and Results Definition of the Pinning Model Assumptions Results 3 Proofs A sketch of the proof Critical Threshold 4 The directed Polymer in a random Environment with Heavy Tails Introduction Improve the Critical threshold 5 Conclusions and open problems N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

4 What is the Pinning Model? Consider a Point Process τ N 0. τ 0 τ 1 τ 2 τ m-1 τ m-1 τ m 0 N N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

5 What is the Pinning Model? Consider a Point Process τ N 0. τ 0 τ 1 τ 2 τ m-1 τ m-1 τ m 0 N Reference example: 0-level set of a Markov Chain S. S 10 S 12 S 1 S 7 S 9 S 13 S 17 S 18 S 11 S 2 S 4 S14 S 0 S 6 S 8 S 16 S 3 S 5 S 15 τ 0 τ 1 τ 2 τ 3 τ 4 τ 5 τ 6 τ 7 = It is useful thought (τ, P) as a random subset of N. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

6 What is the Pinning Model? Consider a Point Process τ N 0. τ 0 τ 1 τ 2 τ m-1 τ m-1 τ m 0 N Fix N and put a weight ω n on each integer smaller than N. τ 0 τ 1 τ 2 τ m-1 τ m-1 τ m... 0 ω 1 ω 2 ω 3 ω 4 ω 5 ω N-5 ω N-4 ω N-3 ω N-2 ω N-1 N The sequence ω = (ω n ) N n=1 is the disorder: at each point s j we give a reward/penalty to visit the point proportional to the value of ω sj. We have an interaction between τ and ω described by a given function Φ ω N ( ). Therefore the Pinning Model is P ω N(τ 1 = s 1, τ 2 = s 2,, τ m = s m ) Φ ω N(s 1,, s m )P(τ 1 = s 1, τ 2 = s 2,, τ m = s m ) N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

7 What do we want to know? We want to understand the behavior of the process τ with respect to the Pinning Model P ω N in the limit N : Will the behavior of (τ, P ω N ) be the same of the original process (τ, P)? In what way will the behavior depend of the disorder? N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

8 What do we want to know? We want to understand the behavior of the process τ with respect to the Pinning Model P ω N in the limit N : Will the behavior of (τ, P ω N ) be the same of the original process (τ, P)? In what way will the behavior depend of the disorder? We will consider the rescaled process τ/n. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

9 2 Definitions and Results Definition of the Pinning Model Assumptions Results

10 Definition of the Pinning Model Let (τ, P) be a Point process starting from 0 consider Let P N be the law of τ/n [0, 1]. τ/n. Consider a probability measure defined by a Radon-Nikodym derivative: I [0, 1] dp ω ( N 1 ) β,h,n (I) exp (βω n h)1 dp n/n I. N n=1 This definition depends of three parameters: h, β and ω = (ω n ) n N a real sequence called disorder. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

11 Definition of the Pinning Model Add a normalization constant to make P a probability: for I [0, 1] dp ω β,h,n (I) = 1 ( N 1 ) 1 I dp N Z ω exp (βω n h)1 n/n I. β,h,n P N be the law of τ/n [0, 1], 1 1 I boundary condition, Z ω β,h,n normalization constant. n=1 N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

12 Definition of the Pinning Model Add a normalization constant to make P a probability: for I [0, 1] dp ω β,h,n (I) = 1 ( N 1 ) 1 I dp N Z ω exp (βω n h)1 n/n I. β,h,n P N be the law of τ/n [0, 1], 1 1 I boundary condition, Z ω β,h,n normalization constant. The Pinning Model is a probability measure on the space of all closed sets of [0, 1] which contain 0, 1. n=1 N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

13 2 Definitions and Results Definition of the Pinning Model Assumptions Results

14 Assumption on the Point Process We consider (τ, P) a Renewal process, that is a Point Process such that τ 0 = 0, (τ j τ j 1 ) j N is an i.i.d. sequence. Let K(n) = P(τ 1 = n), which is a probability on N { }. If K( ) > 0 Def. τ is terminating τ <, P-a.s. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

15 Assumption on the Point Process We consider (τ, P) a Renewal process, that is a Point Process such that τ 0 = 0, (τ j τ j 1 ) j N is an i.i.d. sequence. Let K(n) = P(τ 1 = n), which is a probability on N { }. If K( ) > 0 Def. τ is terminating τ <, P-a.s. We consider a non-terminating renewal process τ s.t. Precisely K(n) = e Cnγ γ (0, 1). 1 subexponential: lim n K(n + k)/k(n) = 1 for any k > 0 and lim n K (2) (n)/k(n) = 2, 2 stretched-exponential: lim n log K(n)/N γ = C, for some C > 0 and γ (0, 1). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

16 Assumption on the Parameters The disorder ω is a fixed (quenched) realization of an i.i.d. sequence of random variables such that P(ω 1 > t) ct α, t. α (0, 1) Moreover we assume ω 1 positives with a continuous distribution. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

17 Assumption on the Parameters The disorder ω is a fixed (quenched) realization of an i.i.d. sequence of random variables such that P(ω 1 > t) ct α, t. α (0, 1) Moreover we assume ω 1 positives with a continuous distribution. β = β N = ˆβN γ 1 α 0 as N. h > 0 fixed. N 1 n=1 (βω n h)1 n/n I N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

18 The role of β and h the role of β is to tune the intensity of the disorder ω. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

19 The role of β and h the role of β is to tune the intensity of the disorder ω. the role of h > 0 is to force away the renewal process: consider the Homogeneous Pinning Model (β = 0), then if I = m P ω 0,h,N(I) = 1 1 I Z ω e ( (m 1)h) m K (N(x i x i 1 )) 0,h,N i=1 = 1 m 1 I e h Z ω e h K (N(x i x i 1 )). 0,h,N i=1 K(n) = e h K(n) = P ω 0,h,N (I) = P cons N (I) : the law of a (rescaled) terminating Renewal Process constrained to visit 1. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

20 The role of β and h the role of β is to tune the intensity of the disorder ω. the role of h > 0 is to force away the renewal process. Homogeneous Pinning Model (β = 0): = P ω 0,h,N (I) = P cons N (I) : the law of a (rescaled) terminating Renewal Process constrained to visit 1. This forces the typical trajectories to have big jumps: h>0 h=0 N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

21 Remark about h > 0 If h > 0, then we can assume to be 0 by replacing the original renewal process with a new terminating one. ( N 1 P ω l β,h,n(i) exp β ω n 1 n/n I) K (N(x i x i 1 )) n=1 i=1 where K is a terminating renewal process. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

22 Remark about h > 0 If h > 0, then we can assume to be 0 by replacing the original renewal process with a new terminating one. ( N 1 P ω l β,h,n(i) exp β ω n 1 n/n I) K (N(x i x i 1 )) n=1 i=1 where K is a terminating renewal process. Keep in mind this formula in the sequel. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

23 2 Definitions and Results Definition of the Pinning Model Assumptions Results

24 Set-up We have a random set τ/n [0, 1], a modification of its law P ω β N,h,N. Goal: study its behavior as N. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

25 Set-up We have a random set τ/n [0, 1], a modification of its law P ω β N,h,N. Goal: study its behavior as N.= Specify the space and the topology. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

26 Set-up We have a random set τ/n [0, 1], a modification of its law P ω β N,h,N. Goal: study its behavior as N.= Specify the space and the topology. We look at τ/n [0, 1] as a random variables on the space of the all closed subsets of [0, 1] which contain 0, 1. We equip this space with the Hausdorff distance: Hausdorff distance: d H (A, B) < ɛ Def. a A, b B : a b < ɛ and vice-versa, interchanging A and B. B A b a ε N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

27 Theorems: Concentration Theorem (Concentration of the Renewal Process) For any N N there exists a set Iβ ω N,N [0, 1] depending only of ω, α and γ s.t. for any fixed ɛ > 0 P ω β N,h,N ( d H (I, I ω β N,N) > ɛ) P 0, Here d H (, ) denotes the Hausdorff distance. ε ε ε ε N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

28 Theorems: Convergence Theorem (Limit Theorem) For any ˆβ > 0 there exists a random closed subset Î [0, 1], function wˆβ, of a suitable continuum disorder w, such that I ω β N,N (d) Îwˆβ, on (X, d H ), the space of all closed sets of [0, 1] which contain 0, 1. The limit set Î depends only of w, α and γ. wˆβ, N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

29 Conclusion: Concentration & Convergence Corollary (Convergence of the Renewal Process) N If β N ˆβN γ 1 α, then τ/n [0, 1] (d) Î wˆβ, in (X, d H ), with respect to the Pinning Model measure P ω β N,h,N. ε ε ε Limit Set Renewal Set Size of the mesh 1/N N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

30 Critical Threshold Theorem (Critical Threshold) Let w the continuum disorder, then there exists a random variable ˆβ w c 1 1 If ˆβ < ˆβ w c then Îwˆβ, {0, 1}, a.e.-w s.t If ˆβ > ˆβ w c then Îwˆβ, {0, 1}, a.e.-w Moreover ˆβ w c > 0 for a.e.-w and for any choice of α, γ (0, 1). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

31 Interpretation of the Critical Threshold If ˆβ < ˆβ c, the disorder is irrelevant : the behavior is like the one of the Homogeneous Pinning Model β = 0. ε ε If ˆβ > ˆβ c different. ε ε ε ε N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

32 3 Proofs A sketch of the proof Critical Threshold

33 Energy-Entropy to visit a set ( N 1 P ω l β,h,n(i) exp β ω n 1 n/n I) K (N(x i x i 1 )) n=1 i=1 Keep in Mind: Any time that we make a jump to visit a point x, we are penalized by e (N( length of the jump))γ, the probability to make the jump, but we gain an Energy given by the disorder on the point x. Case I = {0, x, 1}: we have to make two jumps. Penalization = K(Nx)K(N(1 x)) = e CNγ (x γ +(1 x) γ ) 0 x 1 P ω β,h,n(i) e βωx CNγ (x γ +(1 x) γ ) CN γ (x γ + (1 x) γ ) the Entropy penalization to visit exactly x. ω x the Energy gained by visiting x. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

34 Keep in Mind: Any time that we make a jump to visit a point x, we are penalized by e (N( length of the jump))γ, the probability to make the jump, but we gain an Energy given by the disorder on the point x. i.i.d. structure of the jumps: if I = {x 0 = 0 < x 1 < < x l = 1}, then penalization to visit exactly a fixed set ι is K(Nx 1 )K(N(x 2 x 1 )) K(N(1 x l 1 )) = e CNγ l i=1 (x i x i 1 ) γ ι= { x 0 x 1 x 2 x 3 x 4 x 5 x 6 } x 0 x 1 x 2 x 3 x 4 x 5 x 6 ω 1 ω 2 ω 3 ω 4 ω 5 ω 13 ω 14 ω 15 ω 16 ω 17 l Entropy E N (ι) := CN γ (x i x i 1 ) γ i=1 Energy σ N = l ω xi i=1 ω 6 ω 7 ω 8 ω 9 ω 10 ω 11 ω 12 visit only ι N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

35 The Idea Idea: compare the Energy gained and the Entropy penalization to visit a given set ι.... ω N-5 ω 1 ω 2 ω 3 ω 4 ω 5 ω N-4 ω N-3 ω N-2 ω N-1 The probability to visit (exactly) a set ι P ω β,h,n (ι) βσ N(ι) E N (ι). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

36 The Idea Idea: compare the Energy gained and the Entropy penalization to visit a given set ι.... ω N-5 ω 1 ω 2 ω 3 ω 4 ω 5 ω N-4 ω N-3 ω N-2 ω N-1 Maximize the probability The probability to visit (exactly) a set ι P ω β,h,n (ι) βσ N(ι) E N (ι). I ω β,n = arg max {βσ N (I) E N (I)}. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

37 The Idea Idea: compare the Energy gained and the Entropy penalization to visit a given set ι.... ω N-5 ω 1 ω 2 ω 3 ω 4 ω 5 ω N-4 ω N-3 ω N-2 ω N-1 Maximize the probability The probability to visit (exactly) a set ι P ω β,h,n (ι) βσ N(ι) E N (ι). I ω β,n = arg max {βσ N (I) E N (I)}. Choose β: N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

38 The Idea Idea: compare the Energy gained and the Entropy penalization to visit a given set ι.... ω N-5 ω 1 ω 2 ω 3 ω 4 ω 5 ω N-4 ω N-3 ω N-2 ω N-1 Maximize the probability The probability to visit (exactly) a set ι P ω β,h,n (ι) βσ N(ι) E N (ι). I ω β,n = arg max {βσ N (I) E N (I)}. Choose β: Energy: σ N N 1/α = β Nγ Entropy: E N N γ N 1/α. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

39 The Idea Idea: compare the Energy gained and the Entropy penalization to visit a given set ι.... ω N-5 ω 1 ω 2 ω 3 ω 4 ω 5 ω N-4 ω N-3 ω N-2 ω N-1 Maximize the probability The probability to visit (exactly) a set ι P ω β,h,n (ι) βσ N(ι) E N (ι). I ω β,n = arg max {βσ N (I) E N (I)}. Choose β: Energy: σ N N 1/α = β Nγ Entropy: E N N γ N 1/α. prove concentration around I ω β N,N. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

40 3 Proofs A sketch of the proof Critical Threshold

41 Critical Threshold: the continuum disorder We consider the limit set Î wˆβ,. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

42 Critical Threshold: the continuum disorder We consider the limit set Î wˆβ,. It depends of a continuum disorder w = (M i, Y i ) N. M i = (E E i ) 1/α, α (0, 1), E i E(1) i.i.d. and Y i U([0, 1]) i.i.d. These two sequences are independent. How does it turn out? N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

43 Critical Threshold: the continuum disorder We consider the limit set Î wˆβ,. It depends of a continuum disorder w = (M i, Y i ) N. M i = (E E i ) 1/α, α (0, 1), E i E(1) i.i.d. and Y i U([0, 1]) i.i.d. These two sequences are independent. How does it turn out? We regard the Ordered Statistic of the disorder ω. M (N) i Y (N) i = value of the i-maximum, = position of the i-maximum. ω 1 ω 2 ω 3 ω 4 ω 5 ω 6 ω 7 ω 8 ω 9 ω 10 ω 11 ω 12 ω 13 ω 14 ω 15 ω 16 ω 17 N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

44 Extreme value theory: (N 1/α M (N) i, Y (N) i ) i (d) (M i, Y i ) i N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

45 Critical Threshold: Energy-Entropy balance We define { Î wˆβ, = arg max ˆβσ (I) E(I)}, I σ = σ (M i, Y i ) is the continuum Energy: σ (I) = i:y i I M i E is the continuum Entropy, a suitable extension of the Entropy (if I is finite E coincide with the original Entropy). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

46 Critical Threshold: Positivity ˆβ c = inf{ ˆβ : Î {0, 1}}. wˆβ, For a.e.-w, ˆβ c > 0 for any choice of α, γ (0, 1). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

47 Critical Threshold: Positivity ˆβ c = inf{ ˆβ : Î {0, 1}}. wˆβ, For a.e.-w, ˆβ c > 0 for any choice of α, γ (0, 1). 1 Given ɛ > 0, then Î wˆβ, [0, ɛ] [1 ɛ, 1] for ˆβ small. ε ε N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

48 Critical Threshold: Positivity ˆβ c = inf{ ˆβ : Î {0, 1}}. wˆβ, For a.e.-w, ˆβ c > 0 for any choice of α, γ (0, 1). 1 Given ɛ > 0, then Î wˆβ, [0, ɛ] [1 ɛ, 1] for ˆβ small. ε ε 2 If ɛ > 0 is too small, x ε = total Energy contained in [0,ε] U [1-ε,1] 0 ε 1-ε Energy(Î wˆβ, ) X ɛ ɛ 1/α and Entropy(Î wˆβ, ) Cɛγ + 1. Then ˆβ Energy(Î wˆβ, ) Entropy(Î wˆβ, ) < 1: impossible because Entropy({0, 1}) = 1. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31 1

49 4 The directed Polymer in a random Environment with Heavy Tails Introduction Improve the Critical threshold

50 The directed Polymer 1 Let s be a dimensional simple random walk starting from 0 and constrained to come back to 0 after N-steps. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

51 The directed Polymer 1 Let s be a dimensional simple random walk starting from 0 and constrained to come back to 0 after N-steps. 2 Take an i.i.d. sequence (ω = {ω i,j } i,j N, P) placed on all integers that can be touched by the walk. This sequence is called environment. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

52 Directed Polymer: Gibbs Measure For a given path s we define the Gibbs measure µ β,n (s) = eβσ N(s) Q β,n, where σ N ( ) = i,j ω i,j1 (si =j) is the Energy and Q β,n is a normalization constant. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

53 Directed Polymer: Gibbs Measure For a given path s we define the Gibbs measure µ β,n (s) = eβσ N(s) Q β,n, where σ N ( ) = i,j ω i,j1 (si =j) is the Energy and Q β,n is a normalization constant. A. Auffinger and O. Louidor (2011): the environment has heavy tails with index α (0, 2). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

54 4 The directed Polymer in a random Environment with Heavy Tails Introduction Improve the Critical threshold

55 The Critical threshold Let σ Continuum Energy E Entropy ˆγ β = arg max {βσ (γ) E(γ)} γ N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

56 The Critical threshold Let σ Continuum Energy E Entropy The critical threshold is ˆγ β = arg max {βσ (γ) E(γ)} γ β c = inf{β > 0 : ˆγ β 0}. Theorem (A. Auffinger and O. Louidor, 2011) For a.e. realization of the continuum disorder, β c > 0 for α [0, 1/3) and β c = 0 for [1/2, 2). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

57 The Critical threshold Let σ Continuum Energy E Entropy The critical threshold is ˆγ β = arg max {βσ (γ) E(γ)} γ β c = inf{β > 0 : ˆγ β 0}. Theorem (A. Auffinger and O. Louidor, 2011) For a.e. realization of the continuum disorder, β c > 0 for α [0, 1/3) and β c = 0 for [1/2, 2). AIM: β c > 0 for α [1/3, 1/2). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

58 Step 1 For any ɛ > 0, ˆγ β < ɛ if β is small enough. ε (x,ε) N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

59 Step 2 If ɛ > 0 is small enough, then we cannot never gain enough Energy to compensate any Entropy cost. KEY: Consider the process X ɛ of the Energy in A ɛ A ε ε Conclusion: σ (ˆγ β ) X ɛ ɛ 1/α and E(ˆγ β ) Cɛ 2γ = ˆβσ (ˆγ β ) < E(ˆγ β ). N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

60 Conclusions and open problems 1 γ = 0, that is when K(n) L(n) n 1+γ. Conjecture: β N ˆβN 1/α log N. 2 α > 1 which is linked with the open problem of A. Auffinger and O. Louidor in the case α > 2. N. Torri (Lyon 1 & Milano-Bicocca) Pinning Model Roma, October 8, / 31

61 Thanks for your attention!

62 Entropy Let us consider the space L 0 = {s : [0, 1] R : s is 1 Lipschitz, s(0) = s(1) = 0} equipped with L -norm, denoted by. For a curve γ L 0 we define its Entropy as E(γ) = 1 0 ( ) d e dx γ(x) dx, where e(x) = 1 2 ((1 + x) log(1 + x) + (1 x) log(1 x)).

63 Energy We introduce the continuous environment σ as σ (γ) = i M i δ Zi (graph(γ)), γ L 0. with graph(γ) = {(x, γ(x)) : x [0, 1])} D = {(x, y) R 2 : y x (1 x)} is the graph of γ, M i = (E E i ) 1/α, α (0, 2) and (Z i ) i N is an i.i.d.-sequence of Uniform(D) r.v. s independent of (M i ) i N.

ELEMENTS OF PROBABILITY THEORY

ELEMENTS OF PROBABILITY THEORY ELEMENTS OF PROBABILITY THEORY Elements of Probability Theory A collection of subsets of a set Ω is called a σ algebra if it contains Ω and is closed under the operations of taking complements and countable

More information

Determination of thin elastic inclusions from boundary measurements.

Determination of thin elastic inclusions from boundary measurements. Determination of thin elastic inclusions from boundary measurements. Elena Beretta in collaboration with E. Francini, S. Vessella, E. Kim and J. Lee September 7, 2010 E. Beretta (Università di Roma La

More information

Lower Tail Probabilities and Related Problems

Lower Tail Probabilities and Related Problems Lower Tail Probabilities and Related Problems Qi-Man Shao National University of Singapore and University of Oregon qmshao@darkwing.uoregon.edu . Lower Tail Probabilities Let {X t, t T } be a real valued

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

MATH 564/STAT 555 Applied Stochastic Processes Homework 2, September 18, 2015 Due September 30, 2015

MATH 564/STAT 555 Applied Stochastic Processes Homework 2, September 18, 2015 Due September 30, 2015 ID NAME SCORE MATH 56/STAT 555 Applied Stochastic Processes Homework 2, September 8, 205 Due September 30, 205 The generating function of a sequence a n n 0 is defined as As : a ns n for all s 0 for which

More information

6. Brownian Motion. Q(A) = P [ ω : x(, ω) A )

6. Brownian Motion. Q(A) = P [ ω : x(, ω) A ) 6. Brownian Motion. stochastic process can be thought of in one of many equivalent ways. We can begin with an underlying probability space (Ω, Σ, P) and a real valued stochastic process can be defined

More information

Exercises with solutions (Set D)

Exercises with solutions (Set D) Exercises with solutions Set D. A fair die is rolled at the same time as a fair coin is tossed. Let A be the number on the upper surface of the die and let B describe the outcome of the coin toss, where

More information

Upper and lower bounds for ruin probability

Upper and lower bounds for ruin probability Upper and lower bounds for ruin probability E. Pancheva,Z.Volkovich and L.Morozensky 3 Institute of Mathematics and Informatics, the Bulgarian Academy of Sciences, 3 Sofia, Bulgaria pancheva@math.bas.bg

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

Isodiametric problem in Carnot groups

Isodiametric problem in Carnot groups Conference Geometric Measure Theory Université Paris Diderot, 12th-14th September 2012 Isodiametric inequality in R n Isodiametric inequality: where ω n = L n (B(0, 1)). L n (A) 2 n ω n (diam A) n Isodiametric

More information

Limit theorems for dependent regularly varying functions of Markov chains

Limit theorems for dependent regularly varying functions of Markov chains Limit theorems for functions of with extremal linear behavior Limit theorems for dependent regularly varying functions of In collaboration with T. Mikosch Olivier Wintenberger wintenberger@ceremade.dauphine.fr

More information

arxiv: v1 [math.pr] 11 Dec 2017

arxiv: v1 [math.pr] 11 Dec 2017 Local limits of spatial Gibbs random graphs Eric Ossami Endo Department of Applied Mathematics eric@ime.usp.br Institute of Mathematics and Statistics - IME USP - University of São Paulo Johann Bernoulli

More information

8.1 Concentration inequality for Gaussian random matrix (cont d)

8.1 Concentration inequality for Gaussian random matrix (cont d) MGMT 69: Topics in High-dimensional Data Analysis Falll 26 Lecture 8: Spectral clustering and Laplacian matrices Lecturer: Jiaming Xu Scribe: Hyun-Ju Oh and Taotao He, October 4, 26 Outline Concentration

More information

Function Practice. 1. (a) attempt to form composite (M1) (c) METHOD 1 valid approach. e.g. g 1 (5), 2, f (5) f (2) = 3 A1 N2 2

Function Practice. 1. (a) attempt to form composite (M1) (c) METHOD 1 valid approach. e.g. g 1 (5), 2, f (5) f (2) = 3 A1 N2 2 1. (a) attempt to form composite e.g. ( ) 3 g 7 x, 7 x + (g f)(x) = 10 x N (b) g 1 (x) = x 3 N1 1 (c) METHOD 1 valid approach e.g. g 1 (5),, f (5) f () = 3 N METHOD attempt to form composite of f and g

More information

Random Polymer Models

Random Polymer Models Random Polymer Models Disorder and Localization Phenomena Giambattista Giacomin University of Paris 7 Denis Diderot + Probability Lab. (LPMA) Paris 6 & 7 CNRS U.M.A. 7599 Stochastic Processes in Mathematical

More information

8. Statistical Equilibrium and Classification of States: Discrete Time Markov Chains

8. Statistical Equilibrium and Classification of States: Discrete Time Markov Chains 8. Statistical Equilibrium and Classification of States: Discrete Time Markov Chains 8.1 Review 8.2 Statistical Equilibrium 8.3 Two-State Markov Chain 8.4 Existence of P ( ) 8.5 Classification of States

More information

Almost giant clusters for percolation on large trees

Almost giant clusters for percolation on large trees for percolation on large trees Institut für Mathematik Universität Zürich Erdős-Rényi random graph model in supercritical regime G n = complete graph with n vertices Bond percolation with parameter p(n)

More information

Large deviations for random walks under subexponentiality: the big-jump domain

Large deviations for random walks under subexponentiality: the big-jump domain Large deviations under subexponentiality p. Large deviations for random walks under subexponentiality: the big-jump domain Ton Dieker, IBM Watson Research Center joint work with D. Denisov (Heriot-Watt,

More information

Phase Transitions in Physics and Computer Science. Cristopher Moore University of New Mexico and the Santa Fe Institute

Phase Transitions in Physics and Computer Science. Cristopher Moore University of New Mexico and the Santa Fe Institute Phase Transitions in Physics and Computer Science Cristopher Moore University of New Mexico and the Santa Fe Institute Magnetism When cold enough, Iron will stay magnetized, and even magnetize spontaneously

More information

Hill climbing: Simulated annealing and Tabu search

Hill climbing: Simulated annealing and Tabu search Hill climbing: Simulated annealing and Tabu search Heuristic algorithms Giovanni Righini University of Milan Department of Computer Science (Crema) Hill climbing Instead of repeating local search, it is

More information

1. Stochastic Processes and filtrations

1. Stochastic Processes and filtrations 1. Stochastic Processes and 1. Stoch. pr., A stochastic process (X t ) t T is a collection of random variables on (Ω, F) with values in a measurable space (S, S), i.e., for all t, In our case X t : Ω S

More information

Practical conditions on Markov chains for weak convergence of tail empirical processes

Practical conditions on Markov chains for weak convergence of tail empirical processes Practical conditions on Markov chains for weak convergence of tail empirical processes Olivier Wintenberger University of Copenhagen and Paris VI Joint work with Rafa l Kulik and Philippe Soulier Toronto,

More information

Lecture Notes 3 Convergence (Chapter 5)

Lecture Notes 3 Convergence (Chapter 5) Lecture Notes 3 Convergence (Chapter 5) 1 Convergence of Random Variables Let X 1, X 2,... be a sequence of random variables and let X be another random variable. Let F n denote the cdf of X n and let

More information

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS PROBABILITY: LIMIT THEOREMS II, SPRING 15. HOMEWORK PROBLEMS PROF. YURI BAKHTIN Instructions. You are allowed to work on solutions in groups, but you are required to write up solutions on your own. Please

More information

Diophantine approximation of beta expansion in parameter space

Diophantine approximation of beta expansion in parameter space Diophantine approximation of beta expansion in parameter space Jun Wu Huazhong University of Science and Technology Advances on Fractals and Related Fields, France 19-25, September 2015 Outline 1 Background

More information

Logarithmic scaling of planar random walk s local times

Logarithmic scaling of planar random walk s local times Logarithmic scaling of planar random walk s local times Péter Nándori * and Zeyu Shen ** * Department of Mathematics, University of Maryland ** Courant Institute, New York University October 9, 2015 Abstract

More information

Mañé s Conjecture from the control viewpoint

Mañé s Conjecture from the control viewpoint Mañé s Conjecture from the control viewpoint Université de Nice - Sophia Antipolis Setting Let M be a smooth compact manifold of dimension n 2 be fixed. Let H : T M R be a Hamiltonian of class C k, with

More information

ON THE ZERO-ONE LAW AND THE LAW OF LARGE NUMBERS FOR RANDOM WALK IN MIXING RAN- DOM ENVIRONMENT

ON THE ZERO-ONE LAW AND THE LAW OF LARGE NUMBERS FOR RANDOM WALK IN MIXING RAN- DOM ENVIRONMENT Elect. Comm. in Probab. 10 (2005), 36 44 ELECTRONIC COMMUNICATIONS in PROBABILITY ON THE ZERO-ONE LAW AND THE LAW OF LARGE NUMBERS FOR RANDOM WALK IN MIXING RAN- DOM ENVIRONMENT FIRAS RASSOUL AGHA Department

More information

Lecture I: Asymptotics for large GUE random matrices

Lecture I: Asymptotics for large GUE random matrices Lecture I: Asymptotics for large GUE random matrices Steen Thorbjørnsen, University of Aarhus andom Matrices Definition. Let (Ω, F, P) be a probability space and let n be a positive integer. Then a random

More information

Markov Chain Monte Carlo (MCMC)

Markov Chain Monte Carlo (MCMC) Markov Chain Monte Carlo (MCMC Dependent Sampling Suppose we wish to sample from a density π, and we can evaluate π as a function but have no means to directly generate a sample. Rejection sampling can

More information

A construction of strictly ergodic subshifts having entropy dimension (joint with K. K. Park and J. Lee (Ajou Univ.))

A construction of strictly ergodic subshifts having entropy dimension (joint with K. K. Park and J. Lee (Ajou Univ.)) A construction of strictly ergodic subshifts having entropy dimension (joint with K. K. Park and J. Lee (Ajou Univ.)) Uijin Jung Ajou University, Suwon, South Korea Pingree Park Conference, July 5, 204

More information

Functional Limit theorems for the quadratic variation of a continuous time random walk and for certain stochastic integrals

Functional Limit theorems for the quadratic variation of a continuous time random walk and for certain stochastic integrals Functional Limit theorems for the quadratic variation of a continuous time random walk and for certain stochastic integrals Noèlia Viles Cuadros BCAM- Basque Center of Applied Mathematics with Prof. Enrico

More information

4. Conditional risk measures and their robust representation

4. Conditional risk measures and their robust representation 4. Conditional risk measures and their robust representation We consider a discrete-time information structure given by a filtration (F t ) t=0,...,t on our probability space (Ω, F, P ). The time horizon

More information

Lecture 5. If we interpret the index n 0 as time, then a Markov chain simply requires that the future depends only on the present and not on the past.

Lecture 5. If we interpret the index n 0 as time, then a Markov chain simply requires that the future depends only on the present and not on the past. 1 Markov chain: definition Lecture 5 Definition 1.1 Markov chain] A sequence of random variables (X n ) n 0 taking values in a measurable state space (S, S) is called a (discrete time) Markov chain, if

More information

Simons Workshop on Approximate Counting, Markov Chains and Phase Transitions: Open Problem Session

Simons Workshop on Approximate Counting, Markov Chains and Phase Transitions: Open Problem Session Simons Workshop on Approximate Counting, Markov Chains and Phase Transitions: Open Problem Session Scribes: Antonio Blanca, Sarah Cannon, Yumeng Zhang February 4th, 06 Yuval Peres: Simple Random Walk on

More information

Large Deviations, Linear Statistics, and Scaling Limits for Mahler Ensemble of Complex Random Polynomials

Large Deviations, Linear Statistics, and Scaling Limits for Mahler Ensemble of Complex Random Polynomials Large Deviations, Linear Statistics, and Scaling Limits for Mahler Ensemble of Complex Random Polynomials Maxim L. Yattselev joint work with Christopher D. Sinclair International Conference on Approximation

More information

Let (Ω, F) be a measureable space. A filtration in discrete time is a sequence of. F s F t

Let (Ω, F) be a measureable space. A filtration in discrete time is a sequence of. F s F t 2.2 Filtrations Let (Ω, F) be a measureable space. A filtration in discrete time is a sequence of σ algebras {F t } such that F t F and F t F t+1 for all t = 0, 1,.... In continuous time, the second condition

More information

The Moment Method; Convex Duality; and Large/Medium/Small Deviations

The Moment Method; Convex Duality; and Large/Medium/Small Deviations Stat 928: Statistical Learning Theory Lecture: 5 The Moment Method; Convex Duality; and Large/Medium/Small Deviations Instructor: Sham Kakade The Exponential Inequality and Convex Duality The exponential

More information

Random geometric analysis of the 2d Ising model

Random geometric analysis of the 2d Ising model Random geometric analysis of the 2d Ising model Hans-Otto Georgii Bologna, February 2001 Plan: Foundations 1. Gibbs measures 2. Stochastic order 3. Percolation The Ising model 4. Random clusters and phase

More information

PHYSICS 653 HOMEWORK 1 P.1. Problems 1: Random walks (additional problems)

PHYSICS 653 HOMEWORK 1 P.1. Problems 1: Random walks (additional problems) PHYSICS 653 HOMEWORK 1 P.1 Problems 1: Random walks (additional problems) 1-2. Generating function (Lec. 1.1/1.2, random walks) Not assigned 27. Note: any corrections noted in 23 have not been made. This

More information

State-dependent Importance Sampling for Rare-event Simulation: An Overview and Recent Advances

State-dependent Importance Sampling for Rare-event Simulation: An Overview and Recent Advances State-dependent Importance Sampling for Rare-event Simulation: An Overview and Recent Advances By Jose Blanchet and Henry Lam Columbia University and Boston University February 7, 2011 Abstract This paper

More information

The Contour Process of Crump-Mode-Jagers Branching Processes

The Contour Process of Crump-Mode-Jagers Branching Processes The Contour Process of Crump-Mode-Jagers Branching Processes Emmanuel Schertzer (LPMA Paris 6), with Florian Simatos (ISAE Toulouse) June 24, 2015 Crump-Mode-Jagers trees Crump Mode Jagers (CMJ) branching

More information

INTRODUCTION TO MARKOV CHAINS AND MARKOV CHAIN MIXING

INTRODUCTION TO MARKOV CHAINS AND MARKOV CHAIN MIXING INTRODUCTION TO MARKOV CHAINS AND MARKOV CHAIN MIXING ERIC SHANG Abstract. This paper provides an introduction to Markov chains and their basic classifications and interesting properties. After establishing

More information

The Convergence Rate for the Normal Approximation of Extreme Sums

The Convergence Rate for the Normal Approximation of Extreme Sums The Convergence Rate for the Normal Approximation of Extreme Sums Yongcheng Qi University of Minnesota Duluth WCNA 2008, Orlando, July 2-9, 2008 This talk is based on a joint work with Professor Shihong

More information

Lipschitz matchbox manifolds

Lipschitz matchbox manifolds Lipschitz matchbox manifolds Steve Hurder University of Illinois at Chicago www.math.uic.edu/ hurder F is a C 1 -foliation of a compact manifold M. Problem: Let L be a complete Riemannian smooth manifold

More information

Lecture 2. We now introduce some fundamental tools in martingale theory, which are useful in controlling the fluctuation of martingales.

Lecture 2. We now introduce some fundamental tools in martingale theory, which are useful in controlling the fluctuation of martingales. Lecture 2 1 Martingales We now introduce some fundamental tools in martingale theory, which are useful in controlling the fluctuation of martingales. 1.1 Doob s inequality We have the following maximal

More information

Mathematical Preliminaries

Mathematical Preliminaries Mathematical Preliminaries Economics 3307 - Intermediate Macroeconomics Aaron Hedlund Baylor University Fall 2013 Econ 3307 (Baylor University) Mathematical Preliminaries Fall 2013 1 / 25 Outline I: Sequences

More information

Quentin Berger 1. Introduction and physical motivations

Quentin Berger 1. Introduction and physical motivations ESAIM: PROCEEDINGS AND SURVEYS, October 205, Vol. 5, p. 74-88 A. Garivier et al, Editors e-mail: quentin.berger@upmc.fr INFLUENCE OF DISORDER FOR THE POLYMER PINNING MODEL Quentin Berger Abstract. When

More information

Lower Tail Probabilities and Normal Comparison Inequalities. In Memory of Wenbo V. Li s Contributions

Lower Tail Probabilities and Normal Comparison Inequalities. In Memory of Wenbo V. Li s Contributions Lower Tail Probabilities and Normal Comparison Inequalities In Memory of Wenbo V. Li s Contributions Qi-Man Shao The Chinese University of Hong Kong Lower Tail Probabilities and Normal Comparison Inequalities

More information

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS

PROBABILITY: LIMIT THEOREMS II, SPRING HOMEWORK PROBLEMS PROBABILITY: LIMIT THEOREMS II, SPRING 218. HOMEWORK PROBLEMS PROF. YURI BAKHTIN Instructions. You are allowed to work on solutions in groups, but you are required to write up solutions on your own. Please

More information

Generalized Fuchsian groups and the p-reduction theory of elements in Hurwitz spaces

Generalized Fuchsian groups and the p-reduction theory of elements in Hurwitz spaces Generalized Fuchsian groups and the p-reduction theory of elements in Hurwitz spaces Thomas Weigel, Università di Milano-Bicocca May 8, 2006 A. Hurwitz, 1859-1919 Typeset by FoilTEX 1 Hurwitz groups Definition.

More information

Nonconcave Penalized Likelihood with A Diverging Number of Parameters

Nonconcave Penalized Likelihood with A Diverging Number of Parameters Nonconcave Penalized Likelihood with A Diverging Number of Parameters Jianqing Fan and Heng Peng Presenter: Jiale Xu March 12, 2010 Jianqing Fan and Heng Peng Presenter: JialeNonconcave Xu () Penalized

More information

Lecture 21 Representations of Martingales

Lecture 21 Representations of Martingales Lecture 21: Representations of Martingales 1 of 11 Course: Theory of Probability II Term: Spring 215 Instructor: Gordan Zitkovic Lecture 21 Representations of Martingales Right-continuous inverses Let

More information

Modeling Real Estate Data using Quantile Regression

Modeling Real Estate Data using Quantile Regression Modeling Real Estate Data using Semiparametric Quantile Regression Department of Statistics University of Innsbruck September 9th, 2011 Overview 1 Application: 2 3 4 Hedonic regression data for house prices

More information

Weak quenched limiting distributions of a one-dimensional random walk in a random environment

Weak quenched limiting distributions of a one-dimensional random walk in a random environment Weak quenched limiting distributions of a one-dimensional random walk in a random environment Jonathon Peterson Cornell University Department of Mathematics Joint work with Gennady Samorodnitsky September

More information

Lecture 19: November 10

Lecture 19: November 10 CS294 Markov Chain Monte Carlo: Foundations & Applications Fall 2009 Lecture 19: November 10 Lecturer: Prof. Alistair Sinclair Scribes: Kevin Dick and Tanya Gordeeva Disclaimer: These notes have not been

More information

Simple Abelian Topological Groups. Luke Dominic Bush Hipwood. Mathematics Institute

Simple Abelian Topological Groups. Luke Dominic Bush Hipwood. Mathematics Institute M A E NS G I T A T MOLEM UNIVERSITAS WARWICENSIS Simple Abelian Topological Groups by Luke Dominic Bush Hipwood supervised by Dr Dmitriy Rumynin 4th Year Project Submitted to The University of Warwick

More information

III. Quantum ergodicity on graphs, perspectives

III. Quantum ergodicity on graphs, perspectives III. Quantum ergodicity on graphs, perspectives Nalini Anantharaman Université de Strasbourg 24 août 2016 Yesterday we focussed on the case of large regular (discrete) graphs. Let G = (V, E) be a (q +

More information

Markov Chains CK eqns Classes Hitting times Rec./trans. Strong Markov Stat. distr. Reversibility * Markov Chains

Markov Chains CK eqns Classes Hitting times Rec./trans. Strong Markov Stat. distr. Reversibility * Markov Chains Markov Chains A random process X is a family {X t : t T } of random variables indexed by some set T. When T = {0, 1, 2,... } one speaks about a discrete-time process, for T = R or T = [0, ) one has a continuous-time

More information

Chapter 2: Markov Chains and Queues in Discrete Time

Chapter 2: Markov Chains and Queues in Discrete Time Chapter 2: Markov Chains and Queues in Discrete Time L. Breuer University of Kent 1 Definition Let X n with n N 0 denote random variables on a discrete space E. The sequence X = (X n : n N 0 ) is called

More information

< k 2n. 2 1 (n 2). + (1 p) s) N (n < 1

< k 2n. 2 1 (n 2). + (1 p) s) N (n < 1 List of Problems jacques@ucsd.edu Those question with a star next to them are considered slightly more challenging. Problems 9, 11, and 19 from the book The probabilistic method, by Alon and Spencer. Question

More information

Linear ODEs. Existence of solutions to linear IVPs. Resolvent matrix. Autonomous linear systems

Linear ODEs. Existence of solutions to linear IVPs. Resolvent matrix. Autonomous linear systems Linear ODEs p. 1 Linear ODEs Existence of solutions to linear IVPs Resolvent matrix Autonomous linear systems Linear ODEs Definition (Linear ODE) A linear ODE is a differential equation taking the form

More information

Scaling exponents for certain 1+1 dimensional directed polymers

Scaling exponents for certain 1+1 dimensional directed polymers Scaling exponents for certain 1+1 dimensional directed polymers Timo Seppäläinen Department of Mathematics University of Wisconsin-Madison 2010 Scaling for a polymer 1/29 1 Introduction 2 KPZ equation

More information

1 Sequences of events and their limits

1 Sequences of events and their limits O.H. Probability II (MATH 2647 M15 1 Sequences of events and their limits 1.1 Monotone sequences of events Sequences of events arise naturally when a probabilistic experiment is repeated many times. For

More information

MATH 56A: STOCHASTIC PROCESSES CHAPTER 2

MATH 56A: STOCHASTIC PROCESSES CHAPTER 2 MATH 56A: STOCHASTIC PROCESSES CHAPTER 2 2. Countable Markov Chains I started Chapter 2 which talks about Markov chains with a countably infinite number of states. I did my favorite example which is on

More information

RENORMALIZATION OF DYSON S VECTOR-VALUED HIERARCHICAL MODEL AT LOW TEMPERATURES

RENORMALIZATION OF DYSON S VECTOR-VALUED HIERARCHICAL MODEL AT LOW TEMPERATURES RENORMALIZATION OF DYSON S VECTOR-VALUED HIERARCHICAL MODEL AT LOW TEMPERATURES P. M. Bleher (1) and P. Major (2) (1) Keldysh Institute of Applied Mathematics of the Soviet Academy of Sciences Moscow (2)

More information

Feshbach-Schur RG for the Anderson Model

Feshbach-Schur RG for the Anderson Model Feshbach-Schur RG for the Anderson Model John Z. Imbrie University of Virginia Isaac Newton Institute October 26, 2018 Overview Consider the localization problem for the Anderson model of a quantum particle

More information

Collapse transition of the interacting prudent walk

Collapse transition of the interacting prudent walk Collapse transition of the interacting prudent walk Niccolò Torri Joint work with Nicolas Pétrélis Marseille, December 9, 2016 1 Interacting self-avoiding random walk 2 Related Models 3 Results 4 Methods

More information

Large deviations and fluctuation exponents for some polymer models. Directed polymer in a random environment. KPZ equation Log-gamma polymer

Large deviations and fluctuation exponents for some polymer models. Directed polymer in a random environment. KPZ equation Log-gamma polymer Large deviations and fluctuation exponents for some polymer models Timo Seppäläinen Department of Mathematics University of Wisconsin-Madison 211 1 Introduction 2 Large deviations 3 Fluctuation exponents

More information

An invariance result for Hammersley s process with sources and sinks

An invariance result for Hammersley s process with sources and sinks An invariance result for Hammersley s process with sources and sinks Piet Groeneboom Delft University of Technology, Vrije Universiteit, Amsterdam, and University of Washington, Seattle March 31, 26 Abstract

More information

Resistance Growth of Branching Random Networks

Resistance Growth of Branching Random Networks Peking University Oct.25, 2018, Chengdu Joint work with Yueyun Hu (U. Paris 13) and Shen Lin (U. Paris 6), supported by NSFC Grant No. 11528101 (2016-2017) for Research Cooperation with Oversea Investigators

More information

Interfaces in Discrete Thin Films

Interfaces in Discrete Thin Films Interfaces in Discrete Thin Films Andrea Braides (Roma Tor Vergata) Fourth workshop on thin structures Naples, September 10, 2016 An (old) general approach di dimension-reduction In the paper B, Fonseca,

More information

A local time scaling exponent for compact metric spaces

A local time scaling exponent for compact metric spaces A local time scaling exponent for compact metric spaces John Dever School of Mathematics Georgia Institute of Technology Fractals 6 @ Cornell, June 15, 2017 Dever (GaTech) Exit time exponent Fractals 6

More information

Solving the Poisson Disorder Problem

Solving the Poisson Disorder Problem Advances in Finance and Stochastics: Essays in Honour of Dieter Sondermann, Springer-Verlag, 22, (295-32) Research Report No. 49, 2, Dept. Theoret. Statist. Aarhus Solving the Poisson Disorder Problem

More information

Refined Bounds on the Empirical Distribution of Good Channel Codes via Concentration Inequalities

Refined Bounds on the Empirical Distribution of Good Channel Codes via Concentration Inequalities Refined Bounds on the Empirical Distribution of Good Channel Codes via Concentration Inequalities Maxim Raginsky and Igal Sason ISIT 2013, Istanbul, Turkey Capacity-Achieving Channel Codes The set-up DMC

More information

An Effective Model of Facets Formation

An Effective Model of Facets Formation An Effective Model of Facets Formation Dima Ioffe 1 Technion April 2015 1 Based on joint works with Senya Shlosman, Fabio Toninelli, Yvan Velenik and Vitali Wachtel Dima Ioffe (Technion ) Microscopic Facets

More information

Asymptotic results for empirical measures of weighted sums of independent random variables

Asymptotic results for empirical measures of weighted sums of independent random variables Asymptotic results for empirical measures of weighted sums of independent random variables B. Bercu and W. Bryc University Bordeaux 1, France Seminario di Probabilità e Statistica Matematica Sapienza Università

More information

Computational statistics

Computational statistics Computational statistics Markov Chain Monte Carlo methods Thierry Denœux March 2017 Thierry Denœux Computational statistics March 2017 1 / 71 Contents of this chapter When a target density f can be evaluated

More information

Change point in trending regression

Change point in trending regression Charles University, Prague ROBUST 2010 joint work with A. Aue, L.Horváth, J. Picek (Charles University, Prague) 1 (Charles University, Prague) 1 2 model-formulation (Charles University, Prague) 1 2 model-formulation

More information

Finite-dimensional spaces. C n is the space of n-tuples x = (x 1,..., x n ) of complex numbers. It is a Hilbert space with the inner product

Finite-dimensional spaces. C n is the space of n-tuples x = (x 1,..., x n ) of complex numbers. It is a Hilbert space with the inner product Chapter 4 Hilbert Spaces 4.1 Inner Product Spaces Inner Product Space. A complex vector space E is called an inner product space (or a pre-hilbert space, or a unitary space) if there is a mapping (, )

More information

Learning discrete graphical models via generalized inverse covariance matrices

Learning discrete graphical models via generalized inverse covariance matrices Learning discrete graphical models via generalized inverse covariance matrices Duzhe Wang, Yiming Lv, Yongjoon Kim, Young Lee Department of Statistics University of Wisconsin-Madison {dwang282, lv23, ykim676,

More information

Inference for High Dimensional Robust Regression

Inference for High Dimensional Robust Regression Department of Statistics UC Berkeley Stanford-Berkeley Joint Colloquium, 2015 Table of Contents 1 Background 2 Main Results 3 OLS: A Motivating Example Table of Contents 1 Background 2 Main Results 3 OLS:

More information

Convex Optimization CMU-10725

Convex Optimization CMU-10725 Convex Optimization CMU-10725 Simulated Annealing Barnabás Póczos & Ryan Tibshirani Andrey Markov Markov Chains 2 Markov Chains Markov chain: Homogen Markov chain: 3 Markov Chains Assume that the state

More information

A GENERAL SMOOTHING INEQUALITY FOR DISORDERED POLYMERS

A GENERAL SMOOTHING INEQUALITY FOR DISORDERED POLYMERS A GEERAL SMOOTHIG IEQUALITY FOR DISORDERED POLYMERS FRACESCO CARAVEA AD FRAK DE HOLLADER Abstract. This note sharpens the smoothing inequality of Giacomin and Toninelli 7], 8] for disordered polymers.

More information

Stochastic Processes. Winter Term Paolo Di Tella Technische Universität Dresden Institut für Stochastik

Stochastic Processes. Winter Term Paolo Di Tella Technische Universität Dresden Institut für Stochastik Stochastic Processes Winter Term 2016-2017 Paolo Di Tella Technische Universität Dresden Institut für Stochastik Contents 1 Preliminaries 5 1.1 Uniform integrability.............................. 5 1.2

More information

The KPZ line ensemble: a marriage of integrability and probability

The KPZ line ensemble: a marriage of integrability and probability The KPZ line ensemble: a marriage of integrability and probability Ivan Corwin Clay Mathematics Institute, Columbia University, MIT Joint work with Alan Hammond [arxiv:1312.2600 math.pr] Introduction to

More information

Extremal process associated with 2D discrete Gaussian Free Field

Extremal process associated with 2D discrete Gaussian Free Field Extremal process associated with 2D discrete Gaussian Free Field Marek Biskup (UCLA) Based on joint work with O. Louidor Plan Prelude about random fields blame Eviatar! DGFF: definitions, level sets, maximum

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 1 : Measure-theoretic foundations I

Notes 1 : Measure-theoretic foundations I Notes 1 : Measure-theoretic foundations I Math 733-734: Theory of Probability Lecturer: Sebastien Roch References: [Wil91, Section 1.0-1.8, 2.1-2.3, 3.1-3.11], [Fel68, Sections 7.2, 8.1, 9.6], [Dur10,

More information

Weak Ergodicity Breaking WCHAOS 2011

Weak Ergodicity Breaking WCHAOS 2011 Weak Ergodicity Breaking Eli Barkai Bar-Ilan University Bel, Burov, Korabel, Margolin, Rebenshtok WCHAOS 211 Outline Single molecule experiments exhibit weak ergodicity breaking. Blinking quantum dots,

More information

P(X 0 = j 0,... X nk = j k )

P(X 0 = j 0,... X nk = j k ) Introduction to Probability Example Sheet 3 - Michaelmas 2006 Michael Tehranchi Problem. Let (X n ) n 0 be a homogeneous Markov chain on S with transition matrix P. Given a k N, let Z n = X kn. Prove that

More information

Beyond the Gaussian universality class

Beyond the Gaussian universality class Beyond the Gaussian universality class MSRI/Evans Talk Ivan Corwin (Courant Institute, NYU) September 13, 2010 Outline Part 1: Random growth models Random deposition, ballistic deposition, corner growth

More information

Asymptotic properties of the maximum likelihood estimator for a ballistic random walk in a random environment

Asymptotic properties of the maximum likelihood estimator for a ballistic random walk in a random environment Asymptotic properties of the maximum likelihood estimator for a ballistic random walk in a random environment Catherine Matias Joint works with F. Comets, M. Falconnet, D.& O. Loukianov Currently: Laboratoire

More information

Independence and chromatic number (and random k-sat): Sparse Case. Dimitris Achlioptas Microsoft

Independence and chromatic number (and random k-sat): Sparse Case. Dimitris Achlioptas Microsoft Independence and chromatic number (and random k-sat): Sparse Case Dimitris Achlioptas Microsoft Random graphs W.h.p.: with probability that tends to 1 as n. Hamiltonian cycle Let τ 2 be the moment all

More information

Local time path integrals and their application to Lévy random walks

Local time path integrals and their application to Lévy random walks Local time path integrals and their application to Lévy random walks Václav Zatloukal (www.zatlovac.eu) Faculty of Nuclear Sciences and Physical Engineering Czech Technical University in Prague talk given

More information

Worksheet 7, Math 10560

Worksheet 7, Math 10560 Worksheet 7, Math 0560 You must show all of your work to receive credit!. Determine whether the following series and sequences converge or diverge, and evaluate if they converge. If they diverge, you must

More information

Brownian motion. Samy Tindel. Purdue University. Probability Theory 2 - MA 539

Brownian motion. Samy Tindel. Purdue University. Probability Theory 2 - MA 539 Brownian motion Samy Tindel Purdue University Probability Theory 2 - MA 539 Mostly taken from Brownian Motion and Stochastic Calculus by I. Karatzas and S. Shreve Samy T. Brownian motion Probability Theory

More information

MET Workshop: Exercises

MET Workshop: Exercises MET Workshop: Exercises Alex Blumenthal and Anthony Quas May 7, 206 Notation. R d is endowed with the standard inner product (, ) and Euclidean norm. M d d (R) denotes the space of n n real matrices. When

More information

Quantitative recurrence for beta expansion. Wang BaoWei

Quantitative recurrence for beta expansion. Wang BaoWei Introduction Further study Quantitative recurrence for beta expansion Huazhong University of Science and Technology July 8 2010 Introduction Further study Contents 1 Introduction Background Beta expansion

More information

Random Walks Conditioned to Stay Positive

Random Walks Conditioned to Stay Positive 1 Random Walks Conditioned to Stay Positive Bob Keener Let S n be a random walk formed by summing i.i.d. integer valued random variables X i, i 1: S n = X 1 + + X n. If the drift EX i is negative, then

More information