We study the relation between the minimal spanning tree (MST) on many random points and the "near-minimal" tree which is optimal subject to the constraint that a proportion 3 of its edges must be different f...
详细信息
We study the relation between the minimal spanning tree (MST) on many random points and the "near-minimal" tree which is optimal subject to the constraint that a proportion 3 of its edges must be different from those of the MST. Heuristics suggest that, regardless of details of the probability model. the ratio of lengths should scale as 1 + Theta (delta(2)). We prove this scaling result in the model of the lattice with random edge-lengths and in the Euclidean model.
We analyze the unit-demand Euclidean vehicle routeing problem, where n customers are modeled as independent, identically distributed uniform points and have unit demand. We show new lower bounds on the optimal cost fo...
详细信息
We analyze the unit-demand Euclidean vehicle routeing problem, where n customers are modeled as independent, identically distributed uniform points and have unit demand. We show new lower bounds on the optimal cost for the metric vehicle routeing problem and analyze them in this setting. We prove that there exists a constant (c) over cap > 0 such that the iterated tour partitioning heuristic given by Haimovich and Rinnooy Kan (1985) is a (2 - (c) over cap)-approximation algorithm with probability arbitrarily close to I as the number of customers goes to infinity. It has been a longstanding open problem as to whether one can improve upon the factor of 2 given by Haimovich and Rinnooy Kan. We also generalize this, and previous results, to the multidepot case.
Exponential tail bounds are derived for solutions of max-recursive equations and for max-recursive random sequences, which typically arise as functionals of recursive structures, of random trees or in recursive algori...
详细信息
Exponential tail bounds are derived for solutions of max-recursive equations and for max-recursive random sequences, which typically arise as functionals of recursive structures, of random trees or in recursive algorithms. In particular they arise in the worst case analysis of divide and conquer algorithms, in parallel search algorithms or in the height of random tree models. For the proof we determine asymptotic bounds for the moments or for the Laplace transforms and apply a characterization of exponential tail bounds due to Kasahara (1978).
We prove convergence in distribution for the profile (the number of nodes at each level), normalized by its mean, of random recursive trees when the limit ratio alpha of the level and the logarithm of tree size lies i...
详细信息
We prove convergence in distribution for the profile (the number of nodes at each level), normalized by its mean, of random recursive trees when the limit ratio alpha of the level and the logarithm of tree size lies in [0, e). Convergence of all moments is shown to hold only for a E [0, 1] (with only convergence of finite moments when alpha is an element of (1, e)). When the limit ratio is 0 or 1 for which the limit laws are both constant, we prove asymptotic normality for alpha = 0 and a "quicksort type" limit law for alpha = 1, the latter case having additionally a small range where there is no fixed limit law. Our tools are based on the contraction method and method of moments. Similar phenomena also hold for other classes of trees;we apply our tools to binary search trees and give a complete characterization of the profile. The profiles of these random trees represent concrete examples for which the range of convergence in distribution differs from that of convergence of all moments.
We prove convergence in distribution for the profile (the number of nodes at each level), normalized by its mean, of random recursive trees when the limit ratio alpha of the level and the logarithm of tree size lies i...
详细信息
We prove convergence in distribution for the profile (the number of nodes at each level), normalized by its mean, of random recursive trees when the limit ratio alpha of the level and the logarithm of tree size lies in [0, e). Convergence of all moments is shown to hold only for a E [0, 1] (with only convergence of finite moments when alpha is an element of (1, e)). When the limit ratio is 0 or 1 for which the limit laws are both constant, we prove asymptotic normality for alpha = 0 and a "quicksort type" limit law for alpha = 1, the latter case having additionally a small range where there is no fixed limit law. Our tools are based on the contraction method and method of moments. Similar phenomena also hold for other classes of trees;we apply our tools to binary search trees and give a complete characterization of the profile. The profiles of these random trees represent concrete examples for which the range of convergence in distribution differs from that of convergence of all moments.
We obtain a central limit theorem for a general class of additive parameters (costs, observables) associated to three standard Euclidean algorithms, with optimal speed of convergence. We also provide very precise asym...
详细信息
We obtain a central limit theorem for a general class of additive parameters (costs, observables) associated to three standard Euclidean algorithms, with optimal speed of convergence. We also provide very precise asymptotic estimates and error terms for the mean and variance of such parameters. For costs that are lattice (including the number of steps), we go further and establish a local limit theorem, with optimal speed of convergence. We view an algorithm as a dynamical system restricted to rational inputs, and combine tools imported from dynamics, such as transfer operators, with various other techniques: Dirichlet series, Perron's formula, quasi-powers theorems, and the saddle-point method. Such dynamical analyses had previously been used to perform the average-case analysis of algorithms. For the present (dynamical) analysis in distribution, we require estimates on transfer operators when a parameter varies along vertical lines in the complex plane. To prove them, we adapt techniques introduced recently by Dolgopyat in the context of continuous-time dynamics (Ann. Math. 147 (1998) 357). (C) 2004 Elsevier Inc. All rights reserved.
In certain problems in a variety of applied probability settings (from probabilistic analysis of algorithms to statistical physics), the central requirement is to solve a recursive distributional equation of the form ...
详细信息
In certain problems in a variety of applied probability settings (from probabilistic analysis of algorithms to statistical physics), the central requirement is to solve a recursive distributional equation of the form X-d = g((xi(i), X-i), i >= 1). Here (xi(i)) and g((.)) are given and the X-i are independent copies of the unknown distribution X. We survey this area, emphasizing examples where the function g((.)) is essentially a "maximum" or "minimum" function. We draw attention to the theoretical question of endogeny: in the associated recursive tree process X-i, are the X-i measurable functions of the innovations process (xi(i))?
We consider open addressing hashing and implement it by using the Robin Hood strategy;that is, in case of collision, the element that has traveled the farthest can stay in the slot. We hash similar to alphan elements ...
详细信息
We consider open addressing hashing and implement it by using the Robin Hood strategy;that is, in case of collision, the element that has traveled the farthest can stay in the slot. We hash similar to alphan elements into a table of size n where each probe is independent and uniformly distributed over the table, and alpha < 1 is a constant. Let M-n be the maximum search time for any of the elements in the table. We show that with probability tending to one, M-n is an element of[log(2) log n + sigma, log(2) log n + tau] for some constants sigma, tau depending upon alpha only. This is an exponential improvement over the maximum search time in case of the standard FCFS (first come first served) collision strategy and virtually matches the performance of multiple-choice hash methods.
We consider cuckoo hashing as proposed by Pagh and Rodler in 2001. We show that the expected construction time of the hash table is O(n) as long as the two open addressing tables are each of size at least (1 + epsilon...
详细信息
We consider cuckoo hashing as proposed by Pagh and Rodler in 2001. We show that the expected construction time of the hash table is O(n) as long as the two open addressing tables are each of size at least (1 + epsilon)n, where epsilon > 0 and n is the number of data points. Slightly improved bounds are obtained for various probabilities and constraints. The analysis rests on simple properties of branching processes. (C) 2003 Elsevier Science B.V. All rights reserved.
A random suffix search tree is a binary search tree constructed for the suffixes X(i) = 0 (.) B(i)B(i+1)B(i+2)... of a sequence B(1), B(2), B(3),... of independent identically distributed random b-ary digits B(j). Let...
详细信息
A random suffix search tree is a binary search tree constructed for the suffixes X(i) = 0 (.) B(i)B(i+1)B(i+2)... of a sequence B(1), B(2), B(3),... of independent identically distributed random b-ary digits B(j). Let D(n) denote the depth of the node for X(n) in this tree when B(1) is uniform on Z(b). We show that for any value of b > 1, ED(n) = 2 log n + O(log(2)log n), just as for the random binary search tree. We also show that D(n)/ED(n) --> 1 in probability. (C) 2003 Wiley Periodicals, Inc.
暂无评论