The degree of a typical vertex in generalized random intersection graph models

Size: px
Start display at page:

Download "The degree of a typical vertex in generalized random intersection graph models"

Transcription

1 Discrete Matheatics The degree of a typical vertex in generalized rando intersection graph odels Jerzy Jaworski a, Michał Karoński a, Dudley Stark b a Departent of Discrete Matheatics, Ada Mickiewicz University, ul. Uultowska 87, Poznan, Poland b School of Matheatical Sciences, Queen Mary, University of London, London E1 4NS, UK Received 17 May 004; received in revised for 6 January 006; accepted 19 May 006 Available online 11 July 006 Abstract In this paper we consider the degree of a typical vertex in two odels of rando intersection graphs introduced in [E. Godehardt, J. Jaworski, Two odels of rando intersection graphs for classification, in: M. Schwaiger, O. Opitz Eds., Exploratory Data Analysis in Epirical Research, Proceedings of the 5th Annual Conference of the Gesellschaft für Klassifikation e.v., University of Munich, March 14 16, 001, Springer, Berlin, Heidelberg, New York, 00, pp ], the active and passive odels. The active odels are those for which vertices are assigned a rando subset of a list of objects and two vertices are ade adjacent when their subsets intersect. We prove sufficient conditions for vertex degree to be asyptotically Poisson as well as closely related necessary conditions. We also consider the passive odel of intersection graphs, in which objects are vertices and two objects are ade adjacent if there is at least one vertex in the corresponding active odel containing both objects. We prove a necessary condition for vertex degree to be asyptotically Poisson for passive intersection graphs. 006 Elsevier B.V. All rights reserved. Keywords: Rando graphs; Rando intersection graphs 1. Introduction Rando intersection graphs are rando graph odels with dependent edges, as opposed to the odel with independent edges introduced by Erdős and Rényi. The first odel of rando intersection graphs was introduced in [3,4] and generalized rando intersection graph odels were introduced in []. In this paper we look at the degree of a typical vertex of the graphs described in []. Our results generalize soe of the results in [5] on the binoial odel [3,4]. In defining the odel of [], it is helpful to first define a related odel of bipartite graphs, which we denote by BG n, n, P, where n and are positive integers and P = P 0,P 1,...,P with = 1 is a probability distribution on the set of integers [] :={0, 1,,...,}. The independent sets of BG n, n, P are denoted by V and W and the cardinalities of V and W are given by V =n and W =, respectively. E-ail address: D.S.Stark@aths.qul.ac.uk D. Stark X/$ - see front atter 006 Elsevier B.V. All rights reserved. doi: /j.disc

2 J. Jaworski et al. / Discrete Matheatics In the odel BG n, n, P a rando graph is chosen according to the following procedure: 1 The vertex degrees of the eleents in the independent set V are deterined by i.i.d. rando variables Xv, v V, with coon probability distribution P = P 0,P 1,...,P. Given the values of Xv, the set of neighbors Γv W of v is deterined by choosing a subset Γv in such way that 1 PΓv = S Xv = k = k for all S [] of cardinality k = Xv. The subsets are chosen independently of one another. Thus, given S [], PΓv = S = P S S and Γv are utually independent rando eleents. This procedure deterines BG n, n, P uniquely and, in particular, the neighbors of vertices in W are given by Γw ={v V : w Γv}. By construction, we have Xv = Γv. Moreover, we let Ywdenote Yw= Γw. Two odels of rando intersection graphs, called active and passive, are defined in []. The active graphs have vertex set V and v 1,v V are adjacent if and only if Γv 1 Γv =. The active rando intersection graph odel is denoted by IG a n P. The passive graphs have vertex set W and w 1,w W are adjacent if and only if Γw 1 Γw =. The passive rando intersection graph odel is denoted by IG p np. The paraeter will be set to = n α for soe real nuber α 0. We will characteristically ignore the integer part notation and write = n α. Note that the probabilities iplicitly depend on n. Active graphs ay be ore interesting with respect to applications than the passive graphs IG p np,as[] shows that they ay be interpreted as odels of classification. The intersection graph odel proposed in [3,4] can be treated as the active odel with P taken as Binoial, p. We will denote this odel by Gn,, p and call it the binoial odel. On the other hand, it was shown in [] that for any n and, and any distribution P the rando variable Yw, w W the degree of a vertex w in BG n, n, P is Binoialn, μ/ distributed for any w W, where μ = EXv, v V. Moreover Yw 1, Y w,...,yw are utually independent if and only if BG n, n, P = BG n, n, Binoial, p = G n,,p. Therefore, when is taken as Binoialk, p, IG a n P is the sae as IG p P n and both odels are equivalent to Gn,, p. Properties such as the eergence of sall subgraphs and contiguity with the classical rando graph odel Gn, p of Erdős and Rényi, for which there are n vertices and edges are present independently and with probability p, were studied for Gn,, p in [1,3] while the degree of a typical vertex was exained and liiting distributions were derived in [5].. The active graphs In this section we are interested in the vertex degree of a typical vertex of a rando intersection graph IG a n P. We let and P depend on n and exaine convergence in distribution of vertex degree as n. For positive integers a, b we let a b denote the falling factorial a b := aa 1 a b + 1 and let a 0 = 1, 0 0 = 1, 0 b = 0. A siple exaple of active rando intersection graphs which ay be analyzed without difficulty is the degenerate case where P d = 1 for soe d = d. This exaple was considered in []. The degree of a fixed vertex in the active odel is then Binoialn 1, 1 d d / d distributed. Hence, if for soe constant c M = n 1 1 d d c 1 d

3 154 J. Jaworski et al. / Discrete Matheatics as n, then the distribution of a typical vertex will be asyptotically Poissonc. Given sequences a n, b n,we use the notation a n b n to denote a n /b n 1asn.Ifd κ infinitely often for soe constant κ > 0, then d d / d e κ and 1 is ipossible. Therefore, we ay assue that d = o, in which case 1 holds if nd / c or d c 1 1/α/. If α=1, then with d = c the liiting distribution is Poissonc. This contrasts with Gn,, p, which has a copound Poisson liiting distribution when α = 1 and p = c/n; see [5]. The question thus arises as to what distributions P give a liiting Poisson distribution for vertex degree as n. A partial answer to this question is given by Theore 1. Theore 1. Let X be a rando variable with the probability distribution P and denote by μ and σ its expected value and standard deviation, respectively. Denote by ξ the degree of a vertex in an active intersection rando graph IG a n P, where = n α, α > 0. Then i if α < 1 and Eξ c, where c is a nonnegative constant, then alost all vertices of IG a n P are isolated; ii if α 1, Eξ c, and σ = oμ, then ξ converges weakly to the Poissonc distribution; iii if α 1, Eξ c, μ = o, σ dμ for a constant d>0, and μ 3 = EX 3 = Oμ 3, then ξ does not have a liiting Poisson distribution. Proof. The proof of the theore is given in a sequence of leas. Part i is proved in Lea. The proofs of parts ii and iii are given in Leas 5 and 6, respectively, using Leas 1, 3, and 4. Consider IG a n P and fix a vertex v V. We define a i to be a i = Pv has degree i in IG a n P, i = 0, 1,...,n 1. Our first lea gives an exact forula for the probability generating function Fx= n 1 i=0 a ix i of vertex degree in IG a n P. Lea 1. For the odel IG a n P, where P = P 0,P 1,...,P, the probability generating function Fx of vertex degree is given by Fx= k j + 1 Moreover, the expected vertex degree is n 1 M := ja j = F 1 = n 1 1 and the variance of the vertex degree is S := n 1 jj 1a j + M M = F 1 + M M = n 1n 1 x n M M. 4

4 J. Jaworski et al. / Discrete Matheatics Proof. The degree distribution of an arbitrary vertex v in the active odel is given by [] P{v has degree i in IG a n P } n 1 = 1 E k X i Ek X n 1 i, i k k i = 0, 1,...,n 1, where we define 0 0 = 1 and where E k denotes the k-factorial oent operator. The probability generating function Fxof ξ is given by and Since Fx= [ 1 E k X x + E ] k X n 1, k k M = Eξ = F 1 = n 1 E ξ = Eξξ 1 = F 1 = n 1n 1 E k X, k 1 E k X. k j k k =, the forulae for Fx, M, and S given in the lea follow iediately. We use a variant of the degenerate case to help show that ost vertices in IG a n P are isolated alost surely if α < 1 and M is bounded above. Lea. Suppose that α < 1 and M is bounded above. Then P 0 1 as n and the nuber of isolated vertices I has expectation EI np 0 n. Proof. Let ρ = 1 P 0. The rando variable ξ is stochastically larger than the vertex degree of a fixed vertex of the degenerate active graph with P given by P 0 = 1 ρ, P 1 = ρ. Fix v V. The nuber Z of non-isolated vertices in V\{v} in BG n, n, P for the degenerate graph is Binoialn 1, ρ. The degenerate graph has vertex degree distribution ξ given by { 1 ρ + ρ PW = 0 if k = 0, P ξ = k = ρpw = k if k>0, where, conditional on the value of Z, W is BinoialZ, 1/ distributed. Thus, we have the lower bound M =Eξ E ξ= ρew = ρ n 1/. Since n/ and M is bounded, ρ 0. Our next result shows that μ = o is necessary for M c when σ = oμ. Lea 3. Suppose that σ = oμ and M c for a constant c. Then μ = o.

5 156 J. Jaworski et al. / Discrete Matheatics Proof. If it is not true that μ = o, then for soe constant κ > 0 it holds that μ κ for infinitely any.for such, we lower bound the forula for M at 3 by M = n 1 1 j<κ / n 1 1 k,j κ / + k,j κ / k,j<κ / e kj / k<κ / k<κ / n 11 e κ /4 PX κ / PX < κ /, where X is a rando variable with distribution given by P. Chebyshev s inequality gives σ PX < κ / μ κ / = o1, hence M n 11 e κ /4 + o1 for satisfying μ κ. Next we will show that under the assuptions σ = Oμ and μ = o, there is a natural asyptotic relationship between, n, μ, and c. The previous lea shows that when σ = oμ the assuption μ = o is unnecessary. Suppose that v 1,v V. Conditional on the event that v 1 has degree k and v has degree j in BG n, n, P, the nuber of vertices in W to / which v 1, v are utually adjacent is hypergeoetrically distributed with probability distribution function k k x j x j, where x represents the nuber of utually adjacent vertices, and ean kj /. The expected nuber of utually adjacent vertices is k j kj / = μ /. It is reasonable to suppose that if μ / is sall, then it will be close to the probability that v 1, v are utually adjacent to a positive nuber of vertices in W. Hence, μ / should approxiately be the edge density in IG a n P. It is therefore reasonable to suppose that μ / should be close to c/n. We will use the asyptotic result of Lea 4 frequently in succeeding proofs, often in the for μ = O /n. Lea 4. Suppose that α > 0, μ = o, and σ = Oμ. Under these assuptions M c if and only if μ c n. Proof. First suppose that M c. We will find upper and lower bounds on μ / of the for c/n + on 1. Let μ be the second oent μ := j. Because e jk/ 1 jk + jk, 5

6 J. Jaworski et al. / Discrete Matheatics iplies fro which 1 M n 1 = μ 1 jk jk + = 1 μ + μ, M n 1 + μ. Since σ = Oμ we have μ = Oμ, and now μ = o and M = c + o1 give μ c n + on 1. 6 For the lower bound on μ / we use the following inequality which holds as long as j + k<: 1 k j exp = exp j jk j1 k/ j jk j k jk/ 1 1 j + k/. 7 Fix any ε satisfying 0 < ε < inα, 1/. We find that 1 M n 1 j,k n ε j j,k n ε j n ε PX n ε jk/ 1 1 j + k/ j j n ε μ 1 n ε. 1 1 n ε The upper bound 6 iplies μ = O /n, which, together with ε < 1, produces the estiate μ = on ε.asa result, Chebyshev s inequality, 6, and the assuption σ = Oμ and ε < α/ iply PX > n ε = Oσ / n ε = Oμ / n ε = On ε 1 / = on 1. 8

7 158 J. Jaworski et al. / Discrete Matheatics Now, using M = c + o1, we infer that μ c/n + on 1. 9 Cobining 6 and 9 results in μ / c/n. Supposing now that μ / c/n, we will prove that M = c + o1. In light of 5 and the assuptions of the lea, 1 μ / + μ / which, when substituted in 3, iplies = 1 c/n + on 1, li inf M c. 10 n Siilarly, using 7 and again choosing ε satisfying 0 < ε < inα, 1/, we have k,j n ε j j n ε Using 8 and 3, we conclude as a result that μ /1 + On ε = PX n ε c/n + on 1. li sup M c. 11 n The asyptotics 10 and 11 iply M = c + o1. There are choices of P for which M c but μ / c/n fails to hold. Consider the active graphs with P 0 =1 ρ, P = ρ, and ρ = c/n 1 and let W be a Binoialn 1, ρ distributed rando variable. The degree distribution ξ of a typical vertex has distribution { 1 ρ + ρpw = 0 if k = 0, Pξ = k = ρpw = k if k>0. Despite the fact that M = ρ n 1 = c,wehaveμ = ρ, soμ / c/n. In the next lea we give sufficient conditions for ξ to have a liiting Poisson distribution. Lea 5. Suppose that α > 0, σ = oμ, and M c. Under these assuptions the distribution of ξ converges weakly to the Poissonc distribution. Proof. We will deonstrate that li Fx= e c+cx 1 n when x [0, 1], where Fxwas given by. It then follows that the Laplace transfor of ξ converges to the Laplace transfor of a Poisson distributed variable and that the distribution of ξ is asyptotically Poisson. First note that the expression in square brackets in is bounded by 1 whenever x [0, 1]. If we define ω to be ω := in μ/σ, log n, 13

8 then it follows that we have Fx= k j + 1 k μ ωσ j where Chebyshev s inequality iplies Δ := k j + 1 k μ >ωσ P X μ > ωσ = Oω J. Jaworski et al. / Discrete Matheatics x x n 1 n 1 + Δ, 14 = o1. 15 The last equality is a consequence of our assuption σ = oμ and definition 13. We will show that for all k in {k [0,]: k μ ωσ} = 1 c n + on 1, 16 with the on 1 ter bounded uniforly. The asyptotic 1 will then follow fro The analysis showing 16 is siilar to the arguent proving Lea 4. Using 5, we have 1 jk jk + = 1 μk + μ k. For all k such that k μ ωσ,wehave 1 μμ ωσ + μ μ + ωσ 1 μ + μωσ + μ μ + σ log n. 17 By definition 13, ω μ/σ, fro which μ μωσ = o. 18 The last ter in 17 is bounded by σ = oμ and μ = Oμ, giving μ μ + σ log n μ 4 log n = O, Leas 3 and 4 iply μ / c/n, hence 1 c n + on 1.

9 160 J. Jaworski et al. / Discrete Matheatics In a siilar way, choosing ε in the range 0 < ε < inα, 1/, for all k such that k μ ωσ observing that k =Oμ, Lea 4 and the choice of ε iply k/ = On 1/ = on ε, for large enough n 7 gives us j n ε PX n ε 1 PX > n ε [ ] jk 1 1 n ε μk 1 n ε μμ + ωσ 1 n ε. Our assuptions iply 8 and Lea 4, which, together with 18, give us 1 μ + on 1 = 1 c n + on 1. Although α < 1 is not officially excluded in Lea 5, when α < 1 Lea 5 cannot apply because of Lea. For a illustrative exaple, fix α < 1 and consider P 0 = 1 ρ, P 1 = ρ with ρ = μ c/n = o1, in which case σ = ρ ρ?μ. The next lea shows that when σ is of the sae order as μ and assuptions are ade on the third oent of P we never get Poisson convergence. The ean M and variance S of the vertex degree were given in Lea 4. The third oent of P is defined to be μ 3 := k 3. Lea 6. Suppose that α 1, M c, μ = o, and σ dμ for a constant d>0, and that μ 3 = Oμ 3. Then S c + c d when α > 1 and c + c d li inf n S li sup S = O1 n when α = 1. For α 1, it follows that ξ does not have liiting Poisson distribution. Proof. The forula for F 1 obtained fro differentiating twice is F 1 = n 1n 1.

10 J. Jaworski et al. / Discrete Matheatics If we prove for all α 1 that li inf n F 1 d c + c, 19 then it will follow that li inf n S = li inf n F 1 + M M = li inf n F 1 + c c c + c d, resulting in the lower bounds in the stateent of the theore. In the proof of Lea 4 it was convenient to use n ε as a dividing point to use different analyses. In this proof it sees better to use /μ log n for the estiates in 5 below. We begin with Note that F 1 n 1n 0 k /μ log n For all k in the range 0 k /μ log n we ay use the bound μ = Oμ to conclude that there is a constant C>0 such that for all n kμ μ C log n. 0 Thus, we have [ 1 jk ] jk + = 1 μk + μ k = 1 μk 1 μ k μ 1 μk 1 C log n. 1 We conclude fro 5 and 1 that for n large enough, uniforly for all 0 k /μ log n it holds that [ 1 jk ] jk + 1,

11 16 J. Jaworski et al. / Discrete Matheatics and so for large enough n F 1 n 1n = n 1n = n 1n μ 0 k /μ log n 0 k /μ log n 0 k /μ log n 1 C n 1n μ log n = 1 C n 1n μ log n [ 1 1 jk jk ] + μk μ k k 1 kμ μ 0 k /μ log n μ /μ log n<k where we have used 0 at. Observe that Lea 4 iplies k k, 3 n 1n μ μ n μ σ + μ d + 1 n μ 4 d + 1c. 4 Moreover, Markov s inequality applied to the probability distribution given by k /μ, which has expectation μ 3 /μ = Oμ 3 /μ = Oμ, gives us /μ log n<k k μ μ 3 /μ /μ log n = O μ μ log n μ log n = O = oμ n. 5 Now, 19 results fro 3 5. We now coplete the proof of the lea when α > 1 by deriving the upper bound li sup n F 1 d c + c to atch 19. Fix ε such that 0 < ε < inα 1, 1/. 6 For all 0 k n ε, fro 7 we have 0 j n ε [ ] jk/ 1 1 j + k/ 1.

12 J. Jaworski et al. / Discrete Matheatics It follows that F 1 n 0 k n ε = n n 0 k n ε 0 k n ε 1 0 j n ε [ ] jk/ 1 + n 1 j + k/ n ε <k jk/ 1 j + k/ + 0 j n ε μk/ 1 n ε + n ε <j n ε <j + n n ε <k + n n ε <k. 7 Define τ n to be τ n :=. n ε <k Because μ = on ε, by Chebyshev s inequality and Lea 4 we have σ μ n ε n ε 1 τ n = O n ε = O = O and therefore n τ n = o1 by the choice of ε at 6. We now have F 1 n 0 k n ε μ k / 1 n ε + τ nμk/ 1 n ε + τ n + n τ n n μ μ / 1 n ε + n τ n μ / 1 n ε + n τ n + n τ n = n μ μ + o1 = d + 1c + o1, 8 where 8 was shown at 4. When α = 1 we use an arguent siilar to the one for α > 1, but replacing the function n ε by βn, where β < 1 is a positive constant. Note that = n. We also have μ = O1 because of Lea 4, hence βn<k n = On by

13 164 J. Jaworski et al. / Discrete Matheatics Chebyshev s inequality. As a result, 7 becoes F 1 n jk/n 1 j + k/n + 0 k βn 0 j βn O1n O1n 0 k βn 0 k βn O1μ μ + O1 = O1. 0 j βn βn<j n jk 1 n + O n + O 1 [ kμ n + O 1 n ] + O 1 + n βn<k n By assuption M = O1 and the estiate F 1 = O1 gives S = O1 as well. We have shown for all α 1 that Eξ = S + M is bounded, and so the sequence of rando variables ξ as a function of n is tight. If the distribution of ξ were to converge weakly to the Poissonc distribution, then tightness would iply that S would converge to c, but we have shown that li inf n S >c. 3. The passive graphs The results for passive odel given in [] iply the following forula for the distribution of the vertex degree of a typical vertex w in IG p np. Lea 7. Let ψ be a rando variable with the distribution of the vertex degree of a typical vertex w in IG p np. Then i 1 i P{ψ = i}= 1 1 s μ i s+1 n i s + k kp i+s 1 k. s=0 i+s Moreover, the expected vertex degree is [ n ] kk 1 Eψ = Proof. In [] it was shown that ψ has the distribution given by P{w has degree i in IG p np }=P{ψ = i} i 1 i = 1 s 1 EX i s + EX X n i+s 1. i+s s=0 This iplies the first forula iediately. The forula for the expected value of ψ follows fro the fact that the probability of an edge in IG p np is given see [] by [ P{w k w l is an edge of IG p np }=1 1 E ] n X. The rearks on the relationship between the odels Gn,, p, IG a n P, and IG p P n at the end of Section 1 and the results in [5] show that IG p np has asyptotic Poisson vertex degree when α > 1, n= α, p= c 1+α/,

14 J. Jaworski et al. / Discrete Matheatics and = j p j 1 p j. However, one can easily see that the d-degenerate passive graph IG p np with P d =1 for soe d 3 does not have Poisson vertex degree in the liit for the reason that either the vertex degree is 0 or it is at least d 1. Therefore it sees to be justified that the conditions iplying Poisson convergence of the distribution of ψ ust involve P, as well as higher oents of X. We intend to use Lea 7 to get precise results about vertex degree in IG p np in future research, but in Theore we are able to prove a sufficient condition for ψ to be asyptotically Poisson by a different arguent. Theore. Suppose that probability distribution P satisfies i k=3 k = o/n. ii P c/n for soe constant c>0. Then the distribution of ψ is asyptotically Poissonc. Proof. Let us fix w W. The expected nuber of neighbors of w in the passive odel which are adjacent to a vertex in V of degree of 3 or ore in the graph BG n, n, P is bounded by k n n k = o1, 1 k=3 k k=3 where the su in the first bound is the probability that a particular vertex of V has degree at least 3 and has w and a fixed vertex of W\{w} as a neighbor. We ay ignore vertices in V of degree 0 or 1 because they do not contribute any edges to IG p np. Thus edges in IG p np adjacent to w are alost surely deterined by vertices in V of degree and each such vertex corresponds to a K in IG p np. The expected nuber of these K s for which w is in the vertex set of the K and for which there exist ore than one vertex of V adjacent in BG n, n, P to both vertices in the K is bounded by n 1 P 1 = O1/ = o1. We conclude that alost surely each edge adjacent to w in IG p np is deterined by a unique vertex in V of degree inbg n, n, P. Conditional / on v being of degree in BG n, n, P, the probability that v does not deterine an edge adjacent to w is 1 = 1 /. The probability that v contributes an edge adjacent to w, ignoring the graphs for which degv 3, which we showed are alost surely insignificant, is therefore P /. The choices for the vertices in V are independent, hence the degree of w is distributed asyptotically as Binoialn, P / which converges in distribution to Poissonc. It is easy to check that IG p np satisfies the conditions of Theore when α > 1, n = α, p = c 1+α/, and = j p j 1 p j. References [1] J.A. Fill, E.R. Scheineran, K.B. Singer-Cohen, Rando intersection graphs when = ωn: an equivalence theore relating the evolution of the Gn,, p and Gn, p odels, Rando Structures Algoriths [] E. Godehardt, J. Jaworski, Two odels of rando intersection graphs for classification, in: M. Schwaiger, O. Opitz Eds., Exploratory Data Analysis in Epirical Research, Proceedings of the 5th Annual Conference of the Gesellschaft für Klassifikation e.v., University of Munich, March 14 16, 001, Springer, Berlin, Heidelberg, New York, 00, pp [3] M. Karoński, E.R. Scheineran, K.B. Singer-Cohen, On rando intersection graphs: the subgraph proble, Cobin. Probab. Coput [4] K.B. Singer, Rando intersection graphs, Dissertation, Johns Hopkins University, [5] D. Stark, The vertex degree distribution of rando intersection graphs, Rando Structures Algoriths

Research Article On the Isolated Vertices and Connectivity in Random Intersection Graphs

Research Article On the Isolated Vertices and Connectivity in Random Intersection Graphs International Cobinatorics Volue 2011, Article ID 872703, 9 pages doi:10.1155/2011/872703 Research Article On the Isolated Vertices and Connectivity in Rando Intersection Graphs Yilun Shang Institute for

More information

arxiv: v2 [math.co] 3 Dec 2008

arxiv: v2 [math.co] 3 Dec 2008 arxiv:0805.2814v2 [ath.co] 3 Dec 2008 Connectivity of the Unifor Rando Intersection Graph Sion R. Blacburn and Stefanie Gere Departent of Matheatics Royal Holloway, University of London Egha, Surrey TW20

More information

13.2 Fully Polynomial Randomized Approximation Scheme for Permanent of Random 0-1 Matrices

13.2 Fully Polynomial Randomized Approximation Scheme for Permanent of Random 0-1 Matrices CS71 Randoness & Coputation Spring 018 Instructor: Alistair Sinclair Lecture 13: February 7 Disclaier: These notes have not been subjected to the usual scrutiny accorded to foral publications. They ay

More information

arxiv: v1 [cs.ds] 3 Feb 2014

arxiv: v1 [cs.ds] 3 Feb 2014 arxiv:40.043v [cs.ds] 3 Feb 04 A Bound on the Expected Optiality of Rando Feasible Solutions to Cobinatorial Optiization Probles Evan A. Sultani The Johns Hopins University APL evan@sultani.co http://www.sultani.co/

More information

In this chapter, we consider several graph-theoretic and probabilistic models

In this chapter, we consider several graph-theoretic and probabilistic models THREE ONE GRAPH-THEORETIC AND STATISTICAL MODELS 3.1 INTRODUCTION In this chapter, we consider several graph-theoretic and probabilistic odels for a social network, which we do under different assuptions

More information

Polygonal Designs: Existence and Construction

Polygonal Designs: Existence and Construction Polygonal Designs: Existence and Construction John Hegean Departent of Matheatics, Stanford University, Stanford, CA 9405 Jeff Langford Departent of Matheatics, Drake University, Des Moines, IA 5011 G

More information

Statistics and Probability Letters

Statistics and Probability Letters Statistics and Probability Letters 79 2009 223 233 Contents lists available at ScienceDirect Statistics and Probability Letters journal hoepage: www.elsevier.co/locate/stapro A CLT for a one-diensional

More information

Solutions of some selected problems of Homework 4

Solutions of some selected problems of Homework 4 Solutions of soe selected probles of Hoework 4 Sangchul Lee May 7, 2018 Proble 1 Let there be light A professor has two light bulbs in his garage. When both are burned out, they are replaced, and the next

More information

arxiv: v1 [math.pr] 17 May 2009

arxiv: v1 [math.pr] 17 May 2009 A strong law of large nubers for artingale arrays Yves F. Atchadé arxiv:0905.2761v1 [ath.pr] 17 May 2009 March 2009 Abstract: We prove a artingale triangular array generalization of the Chow-Birnbau- Marshall

More information

3.8 Three Types of Convergence

3.8 Three Types of Convergence 3.8 Three Types of Convergence 3.8 Three Types of Convergence 93 Suppose that we are given a sequence functions {f k } k N on a set X and another function f on X. What does it ean for f k to converge to

More information

A Simple Regression Problem

A Simple Regression Problem A Siple Regression Proble R. M. Castro March 23, 2 In this brief note a siple regression proble will be introduced, illustrating clearly the bias-variance tradeoff. Let Y i f(x i ) + W i, i,..., n, where

More information

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and This article appeared in a ournal published by Elsevier. The attached copy is furnished to the author for internal non-coercial research and education use, including for instruction at the authors institution

More information

Biostatistics Department Technical Report

Biostatistics Department Technical Report Biostatistics Departent Technical Report BST006-00 Estiation of Prevalence by Pool Screening With Equal Sized Pools and a egative Binoial Sapling Model Charles R. Katholi, Ph.D. Eeritus Professor Departent

More information

Metric Entropy of Convex Hulls

Metric Entropy of Convex Hulls Metric Entropy of Convex Hulls Fuchang Gao University of Idaho Abstract Let T be a precopact subset of a Hilbert space. The etric entropy of the convex hull of T is estiated in ters of the etric entropy

More information

M ath. Res. Lett. 15 (2008), no. 2, c International Press 2008 SUM-PRODUCT ESTIMATES VIA DIRECTED EXPANDERS. Van H. Vu. 1.

M ath. Res. Lett. 15 (2008), no. 2, c International Press 2008 SUM-PRODUCT ESTIMATES VIA DIRECTED EXPANDERS. Van H. Vu. 1. M ath. Res. Lett. 15 (2008), no. 2, 375 388 c International Press 2008 SUM-PRODUCT ESTIMATES VIA DIRECTED EXPANDERS Van H. Vu Abstract. Let F q be a finite field of order q and P be a polynoial in F q[x

More information

Sharp threshold functions for random intersection graphs via a coupling method.

Sharp threshold functions for random intersection graphs via a coupling method. Sharp threshold functions for random intersection graphs via a coupling method. Katarzyna Rybarczyk Faculty of Mathematics and Computer Science, Adam Mickiewicz University, 60 769 Poznań, Poland kryba@amu.edu.pl

More information

The Methods of Solution for Constrained Nonlinear Programming

The Methods of Solution for Constrained Nonlinear Programming Research Inventy: International Journal Of Engineering And Science Vol.4, Issue 3(March 2014), PP 01-06 Issn (e): 2278-4721, Issn (p):2319-6483, www.researchinventy.co The Methods of Solution for Constrained

More information

Note on generating all subsets of a finite set with disjoint unions

Note on generating all subsets of a finite set with disjoint unions Note on generating all subsets of a finite set with disjoint unions David Ellis e-ail: dce27@ca.ac.uk Subitted: Dec 2, 2008; Accepted: May 12, 2009; Published: May 20, 2009 Matheatics Subject Classification:

More information

IN modern society that various systems have become more

IN modern society that various systems have become more Developent of Reliability Function in -Coponent Standby Redundant Syste with Priority Based on Maxiu Entropy Principle Ryosuke Hirata, Ikuo Arizono, Ryosuke Toohiro, Satoshi Oigawa, and Yasuhiko Takeoto

More information

List Scheduling and LPT Oliver Braun (09/05/2017)

List Scheduling and LPT Oliver Braun (09/05/2017) List Scheduling and LPT Oliver Braun (09/05/207) We investigate the classical scheduling proble P ax where a set of n independent jobs has to be processed on 2 parallel and identical processors (achines)

More information

Characterization of the Line Complexity of Cellular Automata Generated by Polynomial Transition Rules. Bertrand Stone

Characterization of the Line Complexity of Cellular Automata Generated by Polynomial Transition Rules. Bertrand Stone Characterization of the Line Coplexity of Cellular Autoata Generated by Polynoial Transition Rules Bertrand Stone Abstract Cellular autoata are discrete dynaical systes which consist of changing patterns

More information

Uniform Approximation and Bernstein Polynomials with Coefficients in the Unit Interval

Uniform Approximation and Bernstein Polynomials with Coefficients in the Unit Interval Unifor Approxiation and Bernstein Polynoials with Coefficients in the Unit Interval Weiang Qian and Marc D. Riedel Electrical and Coputer Engineering, University of Minnesota 200 Union St. S.E. Minneapolis,

More information

Poisson processes and their properties

Poisson processes and their properties Poisson processes and their properties Poisson processes. collection {N(t) : t [, )} of rando variable indexed by tie t is called a continuous-tie stochastic process, Furtherore, we call N(t) a Poisson

More information

CSE525: Randomized Algorithms and Probabilistic Analysis May 16, Lecture 13

CSE525: Randomized Algorithms and Probabilistic Analysis May 16, Lecture 13 CSE55: Randoied Algoriths and obabilistic Analysis May 6, Lecture Lecturer: Anna Karlin Scribe: Noah Siegel, Jonathan Shi Rando walks and Markov chains This lecture discusses Markov chains, which capture

More information

Block designs and statistics

Block designs and statistics Bloc designs and statistics Notes for Math 447 May 3, 2011 The ain paraeters of a bloc design are nuber of varieties v, bloc size, nuber of blocs b. A design is built on a set of v eleents. Each eleent

More information

Lower Bounds for Quantized Matrix Completion

Lower Bounds for Quantized Matrix Completion Lower Bounds for Quantized Matrix Copletion Mary Wootters and Yaniv Plan Departent of Matheatics University of Michigan Ann Arbor, MI Eail: wootters, yplan}@uich.edu Mark A. Davenport School of Elec. &

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 11 10/15/2008 ABSTRACT INTEGRATION I

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 11 10/15/2008 ABSTRACT INTEGRATION I MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 11 10/15/2008 ABSTRACT INTEGRATION I Contents 1. Preliinaries 2. The ain result 3. The Rieann integral 4. The integral of a nonnegative

More information

Machine Learning Basics: Estimators, Bias and Variance

Machine Learning Basics: Estimators, Bias and Variance Machine Learning Basics: Estiators, Bias and Variance Sargur N. srihari@cedar.buffalo.edu This is part of lecture slides on Deep Learning: http://www.cedar.buffalo.edu/~srihari/cse676 1 Topics in Basics

More information

Asymptotics of weighted random sums

Asymptotics of weighted random sums Asyptotics of weighted rando sus José Manuel Corcuera, David Nualart, Mark Podolskij arxiv:402.44v [ath.pr] 6 Feb 204 February 7, 204 Abstract In this paper we study the asyptotic behaviour of weighted

More information

Keywords: Estimator, Bias, Mean-squared error, normality, generalized Pareto distribution

Keywords: Estimator, Bias, Mean-squared error, normality, generalized Pareto distribution Testing approxiate norality of an estiator using the estiated MSE and bias with an application to the shape paraeter of the generalized Pareto distribution J. Martin van Zyl Abstract In this work the norality

More information

Some Proofs: This section provides proofs of some theoretical results in section 3.

Some Proofs: This section provides proofs of some theoretical results in section 3. Testing Jups via False Discovery Rate Control Yu-Min Yen. Institute of Econoics, Acadeia Sinica, Taipei, Taiwan. E-ail: YMYEN@econ.sinica.edu.tw. SUPPLEMENTARY MATERIALS Suppleentary Materials contain

More information

Tail estimates for norms of sums of log-concave random vectors

Tail estimates for norms of sums of log-concave random vectors Tail estiates for nors of sus of log-concave rando vectors Rados law Adaczak Rafa l Lata la Alexander E. Litvak Alain Pajor Nicole Toczak-Jaegerann Abstract We establish new tail estiates for order statistics

More information

TEST OF HOMOGENEITY OF PARALLEL SAMPLES FROM LOGNORMAL POPULATIONS WITH UNEQUAL VARIANCES

TEST OF HOMOGENEITY OF PARALLEL SAMPLES FROM LOGNORMAL POPULATIONS WITH UNEQUAL VARIANCES TEST OF HOMOGENEITY OF PARALLEL SAMPLES FROM LOGNORMAL POPULATIONS WITH UNEQUAL VARIANCES S. E. Ahed, R. J. Tokins and A. I. Volodin Departent of Matheatics and Statistics University of Regina Regina,

More information

Quantum algorithms (CO 781, Winter 2008) Prof. Andrew Childs, University of Waterloo LECTURE 15: Unstructured search and spatial search

Quantum algorithms (CO 781, Winter 2008) Prof. Andrew Childs, University of Waterloo LECTURE 15: Unstructured search and spatial search Quantu algoriths (CO 781, Winter 2008) Prof Andrew Childs, University of Waterloo LECTURE 15: Unstructured search and spatial search ow we begin to discuss applications of quantu walks to search algoriths

More information

The Weierstrass Approximation Theorem

The Weierstrass Approximation Theorem 36 The Weierstrass Approxiation Theore Recall that the fundaental idea underlying the construction of the real nubers is approxiation by the sipler rational nubers. Firstly, nubers are often deterined

More information

Necessity of low effective dimension

Necessity of low effective dimension Necessity of low effective diension Art B. Owen Stanford University October 2002, Orig: July 2002 Abstract Practitioners have long noticed that quasi-monte Carlo ethods work very well on functions that

More information

A := A i : {A i } S. is an algebra. The same object is obtained when the union in required to be disjoint.

A := A i : {A i } S. is an algebra. The same object is obtained when the union in required to be disjoint. 59 6. ABSTRACT MEASURE THEORY Having developed the Lebesgue integral with respect to the general easures, we now have a general concept with few specific exaples to actually test it on. Indeed, so far

More information

Computational and Statistical Learning Theory

Computational and Statistical Learning Theory Coputational and Statistical Learning Theory Proble sets 5 and 6 Due: Noveber th Please send your solutions to learning-subissions@ttic.edu Notations/Definitions Recall the definition of saple based Radeacher

More information

The isomorphism problem of Hausdorff measures and Hölder restrictions of functions

The isomorphism problem of Hausdorff measures and Hölder restrictions of functions The isoorphis proble of Hausdorff easures and Hölder restrictions of functions Doctoral thesis András Máthé PhD School of Matheatics Pure Matheatics Progra School Leader: Prof. Miklós Laczkovich Progra

More information

STRONG LAW OF LARGE NUMBERS FOR SCALAR-NORMED SUMS OF ELEMENTS OF REGRESSIVE SEQUENCES OF RANDOM VARIABLES

STRONG LAW OF LARGE NUMBERS FOR SCALAR-NORMED SUMS OF ELEMENTS OF REGRESSIVE SEQUENCES OF RANDOM VARIABLES Annales Univ Sci Budapest, Sect Cop 39 (2013) 365 379 STRONG LAW OF LARGE NUMBERS FOR SCALAR-NORMED SUMS OF ELEMENTS OF REGRESSIVE SEQUENCES OF RANDOM VARIABLES MK Runovska (Kiev, Ukraine) Dedicated to

More information

An Extension to the Tactical Planning Model for a Job Shop: Continuous-Time Control

An Extension to the Tactical Planning Model for a Job Shop: Continuous-Time Control An Extension to the Tactical Planning Model for a Job Shop: Continuous-Tie Control Chee Chong. Teo, Rohit Bhatnagar, and Stephen C. Graves Singapore-MIT Alliance, Nanyang Technological Univ., and Massachusetts

More information

On the Navier Stokes equations

On the Navier Stokes equations On the Navier Stokes equations Daniel Thoas Hayes April 26, 2018 The proble on the existence and soothness of the Navier Stokes equations is resolved. 1. Proble description The Navier Stokes equations

More information

arxiv: v1 [math.nt] 14 Sep 2014

arxiv: v1 [math.nt] 14 Sep 2014 ROTATION REMAINDERS P. JAMESON GRABER, WASHINGTON AND LEE UNIVERSITY 08 arxiv:1409.411v1 [ath.nt] 14 Sep 014 Abstract. We study properties of an array of nubers, called the triangle, in which each row

More information

Bipartite subgraphs and the smallest eigenvalue

Bipartite subgraphs and the smallest eigenvalue Bipartite subgraphs and the sallest eigenvalue Noga Alon Benny Sudaov Abstract Two results dealing with the relation between the sallest eigenvalue of a graph and its bipartite subgraphs are obtained.

More information

Infinitely Many Trees Have Non-Sperner Subtree Poset

Infinitely Many Trees Have Non-Sperner Subtree Poset Order (2007 24:133 138 DOI 10.1007/s11083-007-9064-2 Infinitely Many Trees Have Non-Sperner Subtree Poset Andrew Vince Hua Wang Received: 3 April 2007 / Accepted: 25 August 2007 / Published online: 2 October

More information

Numerically repeated support splitting and merging phenomena in a porous media equation with strong absorption. Kenji Tomoeda

Numerically repeated support splitting and merging phenomena in a porous media equation with strong absorption. Kenji Tomoeda Journal of Math-for-Industry, Vol. 3 (C-), pp. Nuerically repeated support splitting and erging phenoena in a porous edia equation with strong absorption To the eory of y friend Professor Nakaki. Kenji

More information

Lost-Sales Problems with Stochastic Lead Times: Convexity Results for Base-Stock Policies

Lost-Sales Problems with Stochastic Lead Times: Convexity Results for Base-Stock Policies OPERATIONS RESEARCH Vol. 52, No. 5, Septeber October 2004, pp. 795 803 issn 0030-364X eissn 1526-5463 04 5205 0795 infors doi 10.1287/opre.1040.0130 2004 INFORMS TECHNICAL NOTE Lost-Sales Probles with

More information

INDEPENDENT SETS IN HYPERGRAPHS

INDEPENDENT SETS IN HYPERGRAPHS INDEPENDENT SETS IN HYPERGRAPHS Abstract. Many iportant theores and conjectures in cobinatorics, such as the theore of Szeerédi on arithetic progressions and the Erdős-Stone Theore in extreal graph theory,

More information

Computable Shell Decomposition Bounds

Computable Shell Decomposition Bounds Coputable Shell Decoposition Bounds John Langford TTI-Chicago jcl@cs.cu.edu David McAllester TTI-Chicago dac@autoreason.co Editor: Leslie Pack Kaelbling and David Cohn Abstract Haussler, Kearns, Seung

More information

A Bernstein-Markov Theorem for Normed Spaces

A Bernstein-Markov Theorem for Normed Spaces A Bernstein-Markov Theore for Nored Spaces Lawrence A. Harris Departent of Matheatics, University of Kentucky Lexington, Kentucky 40506-0027 Abstract Let X and Y be real nored linear spaces and let φ :

More information

Estimating Entropy and Entropy Norm on Data Streams

Estimating Entropy and Entropy Norm on Data Streams Estiating Entropy and Entropy Nor on Data Streas Ait Chakrabarti 1, Khanh Do Ba 1, and S. Muthukrishnan 2 1 Departent of Coputer Science, Dartouth College, Hanover, NH 03755, USA 2 Departent of Coputer

More information

Closed-form evaluations of Fibonacci Lucas reciprocal sums with three factors

Closed-form evaluations of Fibonacci Lucas reciprocal sums with three factors Notes on Nuber Theory Discrete Matheatics Print ISSN 30-32 Online ISSN 2367-827 Vol. 23 207 No. 2 04 6 Closed-for evaluations of Fibonacci Lucas reciprocal sus with three factors Robert Frontczak Lesbank

More information

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation Course Notes for EE7C (Spring 018: Convex Optiization and Approxiation Instructor: Moritz Hardt Eail: hardt+ee7c@berkeley.edu Graduate Instructor: Max Sichowitz Eail: sichow+ee7c@berkeley.edu October 15,

More information

Probability Distributions

Probability Distributions Probability Distributions In Chapter, we ephasized the central role played by probability theory in the solution of pattern recognition probles. We turn now to an exploration of soe particular exaples

More information

APPROXIMATION BY MODIFIED SZÁSZ-MIRAKYAN OPERATORS

APPROXIMATION BY MODIFIED SZÁSZ-MIRAKYAN OPERATORS APPROXIMATION BY MODIFIED SZÁSZ-MIRAKYAN OPERATORS Received: 23 Deceber, 2008 Accepted: 28 May, 2009 Counicated by: L. REMPULSKA AND S. GRACZYK Institute of Matheatics Poznan University of Technology ul.

More information

Prerequisites. We recall: Theorem 2 A subset of a countably innite set is countable.

Prerequisites. We recall: Theorem 2 A subset of a countably innite set is countable. Prerequisites 1 Set Theory We recall the basic facts about countable and uncountable sets, union and intersection of sets and iages and preiages of functions. 1.1 Countable and uncountable sets We can

More information

Soft Computing Techniques Help Assign Weights to Different Factors in Vulnerability Analysis

Soft Computing Techniques Help Assign Weights to Different Factors in Vulnerability Analysis Soft Coputing Techniques Help Assign Weights to Different Factors in Vulnerability Analysis Beverly Rivera 1,2, Irbis Gallegos 1, and Vladik Kreinovich 2 1 Regional Cyber and Energy Security Center RCES

More information

Testing Properties of Collections of Distributions

Testing Properties of Collections of Distributions Testing Properties of Collections of Distributions Reut Levi Dana Ron Ronitt Rubinfeld April 9, 0 Abstract We propose a fraework for studying property testing of collections of distributions, where the

More information

The Distribution of the Covariance Matrix for a Subset of Elliptical Distributions with Extension to Two Kurtosis Parameters

The Distribution of the Covariance Matrix for a Subset of Elliptical Distributions with Extension to Two Kurtosis Parameters journal of ultivariate analysis 58, 96106 (1996) article no. 0041 The Distribution of the Covariance Matrix for a Subset of Elliptical Distributions with Extension to Two Kurtosis Paraeters H. S. Steyn

More information

Model Fitting. CURM Background Material, Fall 2014 Dr. Doreen De Leon

Model Fitting. CURM Background Material, Fall 2014 Dr. Doreen De Leon Model Fitting CURM Background Material, Fall 014 Dr. Doreen De Leon 1 Introduction Given a set of data points, we often want to fit a selected odel or type to the data (e.g., we suspect an exponential

More information

Analysis of Polynomial & Rational Functions ( summary )

Analysis of Polynomial & Rational Functions ( summary ) Analysis of Polynoial & Rational Functions ( suary ) The standard for of a polynoial function is ( ) where each of the nubers are called the coefficients. The polynoial of is said to have degree n, where

More information

Research Article Some Formulae of Products of the Apostol-Bernoulli and Apostol-Euler Polynomials

Research Article Some Formulae of Products of the Apostol-Bernoulli and Apostol-Euler Polynomials Discrete Dynaics in Nature and Society Volue 202, Article ID 927953, pages doi:055/202/927953 Research Article Soe Forulae of Products of the Apostol-Bernoulli and Apostol-Euler Polynoials Yuan He and

More information

A Note on Online Scheduling for Jobs with Arbitrary Release Times

A Note on Online Scheduling for Jobs with Arbitrary Release Times A Note on Online Scheduling for Jobs with Arbitrary Release Ties Jihuan Ding, and Guochuan Zhang College of Operations Research and Manageent Science, Qufu Noral University, Rizhao 7686, China dingjihuan@hotail.co

More information

The proofs of Theorem 1-3 are along the lines of Wied and Galeano (2013).

The proofs of Theorem 1-3 are along the lines of Wied and Galeano (2013). A Appendix: Proofs The proofs of Theore 1-3 are along the lines of Wied and Galeano (2013) Proof of Theore 1 Let D[d 1, d 2 ] be the space of càdlàg functions on the interval [d 1, d 2 ] equipped with

More information

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

Probability and Stochastic Processes: A Friendly Introduction for Electrical and Computer Engineers Roy D. Yates and David J. Probability and Stochastic Processes: A Friendly Introduction for Electrical and oputer Engineers Roy D. Yates and David J. Goodan Proble Solutions : Yates and Goodan,1..3 1.3.1 1.4.6 1.4.7 1.4.8 1..6

More information

Computational and Statistical Learning Theory

Computational and Statistical Learning Theory Coputational and Statistical Learning Theory TTIC 31120 Prof. Nati Srebro Lecture 2: PAC Learning and VC Theory I Fro Adversarial Online to Statistical Three reasons to ove fro worst-case deterinistic

More information

1 Proof of learning bounds

1 Proof of learning bounds COS 511: Theoretical Machine Learning Lecturer: Rob Schapire Lecture #4 Scribe: Akshay Mittal February 13, 2013 1 Proof of learning bounds For intuition of the following theore, suppose there exists a

More information

LORENTZ SPACES AND REAL INTERPOLATION THE KEEL-TAO APPROACH

LORENTZ SPACES AND REAL INTERPOLATION THE KEEL-TAO APPROACH LORENTZ SPACES AND REAL INTERPOLATION THE KEEL-TAO APPROACH GUILLERMO REY. Introduction If an operator T is bounded on two Lebesgue spaces, the theory of coplex interpolation allows us to deduce the boundedness

More information

Estimating Parameters for a Gaussian pdf

Estimating Parameters for a Gaussian pdf Pattern Recognition and achine Learning Jaes L. Crowley ENSIAG 3 IS First Seester 00/0 Lesson 5 7 Noveber 00 Contents Estiating Paraeters for a Gaussian pdf Notation... The Pattern Recognition Proble...3

More information

Bootstrapping Dependent Data

Bootstrapping Dependent Data Bootstrapping Dependent Data One of the key issues confronting bootstrap resapling approxiations is how to deal with dependent data. Consider a sequence fx t g n t= of dependent rando variables. Clearly

More information

Distributional transformations, orthogonal polynomials, and Stein characterizations

Distributional transformations, orthogonal polynomials, and Stein characterizations Distributional transforations, orthogonal polynoials, and Stein characterizations LARRY GOLDSTEIN 13 and GESINE REINERT 23 Abstract A new class of distributional transforations is introduced, characterized

More information

Shannon Sampling II. Connections to Learning Theory

Shannon Sampling II. Connections to Learning Theory Shannon Sapling II Connections to Learning heory Steve Sale oyota echnological Institute at Chicago 147 East 60th Street, Chicago, IL 60637, USA E-ail: sale@athberkeleyedu Ding-Xuan Zhou Departent of Matheatics,

More information

Supplement to: Subsampling Methods for Persistent Homology

Supplement to: Subsampling Methods for Persistent Homology Suppleent to: Subsapling Methods for Persistent Hoology A. Technical results In this section, we present soe technical results that will be used to prove the ain theores. First, we expand the notation

More information

Generalized eigenfunctions and a Borel Theorem on the Sierpinski Gasket.

Generalized eigenfunctions and a Borel Theorem on the Sierpinski Gasket. Generalized eigenfunctions and a Borel Theore on the Sierpinski Gasket. Kasso A. Okoudjou, Luke G. Rogers, and Robert S. Strichartz May 26, 2006 1 Introduction There is a well developed theory (see [5,

More information

On Process Complexity

On Process Complexity On Process Coplexity Ada R. Day School of Matheatics, Statistics and Coputer Science, Victoria University of Wellington, PO Box 600, Wellington 6140, New Zealand, Eail: ada.day@cs.vuw.ac.nz Abstract Process

More information

Lecture 3: October 2, 2017

Lecture 3: October 2, 2017 Inforation and Coding Theory Autun 2017 Lecturer: Madhur Tulsiani Lecture 3: October 2, 2017 1 Shearer s lea and alications In the revious lecture, we saw the following stateent of Shearer s lea. Lea 1.1

More information

Semicircle law for generalized Curie-Weiss matrix ensembles at subcritical temperature

Semicircle law for generalized Curie-Weiss matrix ensembles at subcritical temperature Seicircle law for generalized Curie-Weiss atrix ensebles at subcritical teperature Werner Kirsch Fakultät für Matheatik und Inforatik FernUniversität in Hagen, Gerany Thoas Kriecherbauer Matheatisches

More information

Computable Shell Decomposition Bounds

Computable Shell Decomposition Bounds Journal of Machine Learning Research 5 (2004) 529-547 Subitted 1/03; Revised 8/03; Published 5/04 Coputable Shell Decoposition Bounds John Langford David McAllester Toyota Technology Institute at Chicago

More information

On Conditions for Linearity of Optimal Estimation

On Conditions for Linearity of Optimal Estimation On Conditions for Linearity of Optial Estiation Erah Akyol, Kuar Viswanatha and Kenneth Rose {eakyol, kuar, rose}@ece.ucsb.edu Departent of Electrical and Coputer Engineering University of California at

More information

Some Classical Ergodic Theorems

Some Classical Ergodic Theorems Soe Classical Ergodic Theores Matt Rosenzweig Contents Classical Ergodic Theores. Mean Ergodic Theores........................................2 Maxial Ergodic Theore.....................................

More information

Hypergraphs not containing a tight tree with a bounded trunk II: 3-trees with a trunk of size 2

Hypergraphs not containing a tight tree with a bounded trunk II: 3-trees with a trunk of size 2 Hypergraphs not containing a tight tree with a bounded trunk II: -trees with a trunk of size Zoltán Füredi Tao Jiang Alexandr Kostochka Dhruv Mubayi Jacques Verstraëte July 6, 08 Abstract A tight r-tree

More information

THE AVERAGE NORM OF POLYNOMIALS OF FIXED HEIGHT

THE AVERAGE NORM OF POLYNOMIALS OF FIXED HEIGHT THE AVERAGE NORM OF POLYNOMIALS OF FIXED HEIGHT PETER BORWEIN AND KWOK-KWONG STEPHEN CHOI Abstract. Let n be any integer and ( n ) X F n : a i z i : a i, ± i be the set of all polynoials of height and

More information

Sampling How Big a Sample?

Sampling How Big a Sample? C. G. G. Aitken, 1 Ph.D. Sapling How Big a Saple? REFERENCE: Aitken CGG. Sapling how big a saple? J Forensic Sci 1999;44(4):750 760. ABSTRACT: It is thought that, in a consignent of discrete units, a certain

More information

On Certain C-Test Words for Free Groups

On Certain C-Test Words for Free Groups Journal of Algebra 247, 509 540 2002 doi:10.1006 jabr.2001.9001, available online at http: www.idealibrary.co on On Certain C-Test Words for Free Groups Donghi Lee Departent of Matheatics, Uni ersity of

More information

Learnability and Stability in the General Learning Setting

Learnability and Stability in the General Learning Setting Learnability and Stability in the General Learning Setting Shai Shalev-Shwartz TTI-Chicago shai@tti-c.org Ohad Shair The Hebrew University ohadsh@cs.huji.ac.il Nathan Srebro TTI-Chicago nati@uchicago.edu

More information

2 Q 10. Likewise, in case of multiple particles, the corresponding density in 2 must be averaged over all

2 Q 10. Likewise, in case of multiple particles, the corresponding density in 2 must be averaged over all Lecture 6 Introduction to kinetic theory of plasa waves Introduction to kinetic theory So far we have been odeling plasa dynaics using fluid equations. The assuption has been that the pressure can be either

More information

Understanding Machine Learning Solution Manual

Understanding Machine Learning Solution Manual Understanding Machine Learning Solution Manual Written by Alon Gonen Edited by Dana Rubinstein Noveber 17, 2014 2 Gentle Start 1. Given S = ((x i, y i )), define the ultivariate polynoial p S (x) = i []:y

More information

Physics 215 Winter The Density Matrix

Physics 215 Winter The Density Matrix Physics 215 Winter 2018 The Density Matrix The quantu space of states is a Hilbert space H. Any state vector ψ H is a pure state. Since any linear cobination of eleents of H are also an eleent of H, it

More information

Non-uniform Berry Esseen Bounds for Weighted U-Statistics and Generalized L-Statistics

Non-uniform Berry Esseen Bounds for Weighted U-Statistics and Generalized L-Statistics Coun Math Stat 0 :5 67 DOI 0.007/s4004-0-009- Non-unifor Berry Esseen Bounds for Weighted U-Statistics and Generalized L-Statistics Haojun Hu Qi-Man Shao Received: 9 August 0 / Accepted: Septeber 0 / Published

More information

ASYMPTOTICS OF MARKOV KERNELS AND THE TAIL CHAIN

ASYMPTOTICS OF MARKOV KERNELS AND THE TAIL CHAIN ASYMPTOTICS OF MARKOV KERNELS AND THE TAIL CHAIN SIDNEY I. RESNICK AND DAVID ZEBER Abstract. An asyptotic odel for extree behavior of certain Markov chains is the tail chain. Generally taking the for of

More information

1. INTRODUCTION AND RESULTS

1. INTRODUCTION AND RESULTS SOME IDENTITIES INVOLVING THE FIBONACCI NUMBERS AND LUCAS NUMBERS Wenpeng Zhang Research Center for Basic Science, Xi an Jiaotong University Xi an Shaanxi, People s Republic of China (Subitted August 00

More information

A Better Algorithm For an Ancient Scheduling Problem. David R. Karger Steven J. Phillips Eric Torng. Department of Computer Science

A Better Algorithm For an Ancient Scheduling Problem. David R. Karger Steven J. Phillips Eric Torng. Department of Computer Science A Better Algorith For an Ancient Scheduling Proble David R. Karger Steven J. Phillips Eric Torng Departent of Coputer Science Stanford University Stanford, CA 9435-4 Abstract One of the oldest and siplest

More information

Pseudo-marginal Metropolis-Hastings: a simple explanation and (partial) review of theory

Pseudo-marginal Metropolis-Hastings: a simple explanation and (partial) review of theory Pseudo-arginal Metropolis-Hastings: a siple explanation and (partial) review of theory Chris Sherlock Motivation Iagine a stochastic process V which arises fro soe distribution with density p(v θ ). Iagine

More information

Math Reviews classifications (2000): Primary 54F05; Secondary 54D20, 54D65

Math Reviews classifications (2000): Primary 54F05; Secondary 54D20, 54D65 The Monotone Lindelöf Property and Separability in Ordered Spaces by H. Bennett, Texas Tech University, Lubbock, TX 79409 D. Lutzer, College of Willia and Mary, Williasburg, VA 23187-8795 M. Matveev, Irvine,

More information

Sharp Time Data Tradeoffs for Linear Inverse Problems

Sharp Time Data Tradeoffs for Linear Inverse Problems Sharp Tie Data Tradeoffs for Linear Inverse Probles Saet Oyak Benjain Recht Mahdi Soltanolkotabi January 016 Abstract In this paper we characterize sharp tie-data tradeoffs for optiization probles used

More information

MANY physical structures can conveniently be modelled

MANY physical structures can conveniently be modelled Proceedings of the World Congress on Engineering Coputer Science 2017 Vol II Roly r-orthogonal (g, f)-factorizations in Networks Sizhong Zhou Abstract Let G (V (G), E(G)) be a graph, where V (G) E(G) denote

More information

DEPARTMENT OF ECONOMETRICS AND BUSINESS STATISTICS

DEPARTMENT OF ECONOMETRICS AND BUSINESS STATISTICS ISSN 1440-771X AUSTRALIA DEPARTMENT OF ECONOMETRICS AND BUSINESS STATISTICS An Iproved Method for Bandwidth Selection When Estiating ROC Curves Peter G Hall and Rob J Hyndan Working Paper 11/00 An iproved

More information

A Note on Scheduling Tall/Small Multiprocessor Tasks with Unit Processing Time to Minimize Maximum Tardiness

A Note on Scheduling Tall/Small Multiprocessor Tasks with Unit Processing Time to Minimize Maximum Tardiness A Note on Scheduling Tall/Sall Multiprocessor Tasks with Unit Processing Tie to Miniize Maxiu Tardiness Philippe Baptiste and Baruch Schieber IBM T.J. Watson Research Center P.O. Box 218, Yorktown Heights,

More information

Universal algorithms for learning theory Part II : piecewise polynomial functions

Universal algorithms for learning theory Part II : piecewise polynomial functions Universal algoriths for learning theory Part II : piecewise polynoial functions Peter Binev, Albert Cohen, Wolfgang Dahen, and Ronald DeVore Deceber 6, 2005 Abstract This paper is concerned with estiating

More information

Linear recurrences and asymptotic behavior of exponential sums of symmetric boolean functions

Linear recurrences and asymptotic behavior of exponential sums of symmetric boolean functions Linear recurrences and asyptotic behavior of exponential sus of syetric boolean functions Francis N. Castro Departent of Matheatics University of Puerto Rico, San Juan, PR 00931 francis.castro@upr.edu

More information