作者:
Lagergren, JKTH
Stockholm Bioinformat Ctr Dept Numer Anal & Comp Sci S-10044 Stockholm Sweden
This paper treats polynomial-time algorithms for reconstruction of phylogenetic trees. The disc-covering method (DCM) presented by Huson et al. (J. Comput. Biol. 6 (3/4) (1999) 369) is a method that boosts the perform...
详细信息
This paper treats polynomial-time algorithms for reconstruction of phylogenetic trees. The disc-covering method (DCM) presented by Huson et al. (J. Comput. Biol. 6 (3/4) (1999) 369) is a method that boosts the performance of phylogenetic tree construction algorithms. Actually, they gave two variations of DCM-Buneman. The first variation was guaranteed to recover the true tree with high probability from polynomial-length sequences (i.e. polynomial in the number of given taxa), but it was not proven to run in polynomial time. The second variation was guaranteed to run in polynomial time. However, it is a heuristic in the sense that it was not proven to recover the true tree with high probability from polynomial-length sequences. We present an improved DCM. The difference between our improved DCM and the heuristic variation of the original DCM is marginal. The main contribution of this paper is the analysis of the algorithm. Our analysis shows that the improved DCM combines the desirable properties of the two variations of the original DCM. That is, it runs in polynomial time and it recovers the true tree with high probability from polynomial-length sequences. Moreover, this is true when the improved DCM is applied to the Neighbor-Joining, the Buneman, as well as the Agarwala algorithm. A key observation for the result of Huson et al. was that threshold graphs of additive distance functions are chordal. We prove a chordal graph theorem concerning minimal triangulations of threshold graphs constructed from distance functions which are close to being additive. This theorem is the key observation behind our improved DCM and it may be interesting in its own right. (C) 2002 Published by Elsevier Science (USA).
We consider the incremental version of the k-Facility Location Problem, which is a common generalization of the facility location and the k-median problems. The objective is to produce an incremental sequence of facil...
详细信息
We consider the incremental version of the k-Facility Location Problem, which is a common generalization of the facility location and the k-median problems. The objective is to produce an incremental sequence of facility sets F-1 subset of F-2 subset of...subset of F-n, where each F-k contains at most k facilities. An incremental facility sequence or an algorithm producing such a sequence is called c-competitive if the cost of each F-k is at most c times the optimum cost of corresponding k-facility location problem, where c is called competitive ratio. In this paper we present two competitive algorithms for this problem. The first algorithm produces competitive ratio 8 alpha, where alpha is the approximation ratio of k-facility location problem. By recently result (Zhang, Theor. Comput. Sci. 384: 126-135, 2007), we obtain the competitive ratio 16 + 8 root 3 + epsilon. The second algorithm has the competitive ratio Delta + 1, where Delta is the ratio between the maximum and minimum nonzero interpoint distances. The latter result has its self interest, specially for the small metric space with Delta <= 8 alpha - 1.
In this paper, we consider the scheduling problem on identical parallel machines, in which jobs are arriving over time and preemption is not allowed. The goal is to minimize the total completion times. According to th...
详细信息
In this paper, we consider the scheduling problem on identical parallel machines, in which jobs are arriving over time and preemption is not allowed. The goal is to minimize the total completion times. According to the idea of the Delayed-SPT algorithm proposed by Hoogeven and Vestjens [Optimal on-line algorithms for single-machine scheduling. In: Proceedings 5th international conference on integer programming and combinatorial optimization (IPCO). Lecture notes in computer science, vol. 1084. Berlin: Springer;1996. p. 404-14], we give an on-line algorithm for the scheduling problem on m identical parallel machines. We show that this algorithm is 2-competitive and the bound is tight. (C) 2008 Elsevier Ltd. All rights reserved.
In this paper the problem of sorting in various models of magnetic bubble memory systems is studied. Three basic parameters are of interest, namely, the number of steps to sort, the number of switches required, and th...
详细信息
In this paper the problem of sorting in various models of magnetic bubble memory systems is studied. Three basic parameters are of interest, namely, the number of steps to sort, the number of switches required, and the number of control states necessary for the switches. Several sorting algorithms are proposed with respective running times essentially n2, n/2, 1/2 n log2 n, 7/2 n, n log2n, respective numbers of switches essentially, 1, n, 2√n, 2√nlog2n , 1og2n, and respective numbers of control states essentially, 3, 2, 1/8 log2n, 1/8 log2n, and 3 log2n.
The results of a Monte-Carlo simulation of algorithms for statistical analysis of information from detectors in radiation monitors, recording the presence of random, short-time, weak radioactive radiation, are present...
详细信息
The results of a Monte-Carlo simulation of algorithms for statistical analysis of information from detectors in radiation monitors, recording the presence of random, short-time, weak radioactive radiation, are presented. The Neyman-Pearson test and likelihood ratio, moving average, digital recursive filter, a priori probability, half-sums, and relative dispersion methods are examined. The results of a comparative analysis of the algorithms and the investigation of the possibilities and conditions for realizing their maximum possibilities are presented. The analysis has shown that the moving average method has advantages. It can be replaced by the similarity ratio method in the event of a Poisson distribution and the a priori probability method in the event of a Gaussian distribution. For an average background of not more than 5 with a large separation the moving average and similarity ratio methods are best. Otherwise the digital recursive filter method is best. The best results are obtained by combining the detection method with the moving average method together with likelihood ratio and a priori probability methods.
The optimization versions of the 3-Partitioning and the Kernel 3-Partitioning problems are considered in this paper. For the objective to maximize the minimum load of the m subsets, it is shown that the MODIFIED LPT a...
详细信息
The optimization versions of the 3-Partitioning and the Kernel 3-Partitioning problems are considered in this paper. For the objective to maximize the minimum load of the m subsets, it is shown that the MODIFIED LPT algorithm has performance ratios (3m - 1)/(4m - 2) and (2m - 1)/(3m - 2), respectively, in the worst case.
A general algorithm, called ALG, for online and semi-online scheduling problem Pm || C-max with m >= 2 is introduced. For the semi-online version, it is supposed that all job have their processing times within the ...
详细信息
A general algorithm, called ALG, for online and semi-online scheduling problem Pm || C-max with m >= 2 is introduced. For the semi-online version, it is supposed that all job have their processing times within the interval [p, rp], where p > 0, 1 < r <= m/m - 1. ALG is a generalization of LS and is optimal in the sense that there is not an algorithm with smaller competitive ratio than that of ALG.
In the classical multiprocessor scheduling problems, it is assumed that the problems are considered in off\|line or on\|line environment. But in practice, problems are often not really off\|line or on\|line but someh...
详细信息
In the classical multiprocessor scheduling problems, it is assumed that the problems are considered in off\|line or on\|line environment. But in practice, problems are often not really off\|line or on\|line but somehow in between. This means that, with respect to the on\|line problem, some further information about the tasks is available, which allows the improvement of the performance of the best possible algorithms. Problems of this class are called semi on\|line ones. The authors studied two semi on\|line multiprocessor scheduling problems, in which, the total processing time of all tasks is known in advance, or all processing times lie in a given interval. They proposed approximation algorithms for minimizing the makespan and analyzed their performance guarantee. The algorithms improve the known results for 3 or more processor cases in the literature.
Fibonacci words are among those words containing a repetition, i.e., at least 2 consecutive equal factors. Main and Lorentz have introduced an algorithm to find one repetition or square in a word. In the word aababb...
详细信息
Fibonacci words are among those words containing a repetition, i.e., at least 2 consecutive equal factors. Main and Lorentz have introduced an algorithm to find one repetition or square in a word. In the word aababba, for example, abab is a repetition or square. An algorithm is presented to compute all the repetitions of factors in a word. The algorithm uses an improved version of the well-known partitioning technique. Proofs are presented to demonstrate that the new algorithm, referred to as REP, is optimal in the class of algorithms computing all the repetitions of a word. Proofs also are presented to show that an associated procedure, REPMAX, is optimal in the class of algorithms computing all the maximal repetitions of a word.
Following the model introduced by Aguech et al. (Probab Eng Inf Sci 21:133-141, 2007), the weighted depth of a node in a labelled rooted tree is the sum of all labels on the path connecting the node to the root. We an...
详细信息
Following the model introduced by Aguech et al. (Probab Eng Inf Sci 21:133-141, 2007), the weighted depth of a node in a labelled rooted tree is the sum of all labels on the path connecting the node to the root. We analyse weighted depths of nodes with given labels, the last inserted node, nodes ordered as visited by the depth first search process, the weighted path length and the weighted Wiener index in a random binary search tree. We establish three regimes of nodes depending on whether the second-order behaviour of their weighted depths follows from fluctuations of the keys on the path, the depth of the nodes or both. Finally, we investigate a random distribution function on the unit interval arising as scaling limit for weighted depths of nodes with at most one child.
暂无评论