On-line Learning and the Metrical Task System Problem

The problem of combining expert advice, studied extensively in the Computational Learning Theory literature, and the Metrical Task System (MTS) problem, studied extensively in the area of On-line Algorithms, contain a number of interesting similarities. In this paper we explore the relationship between these problems and show how algorithms designed for each can be used to achieve good bounds and new approaches for solving the other. Specific contributions of this paper include: An analysis of how two recent algorithms for the MTS problem can be applied to the problem of tracking the best expert in the decision-theoretic setting, providing good bounds and an approach of a much different flavor from the well-known multiplicative-update algorithms. An analysis showing how the standard randomized Weighted Majority (or Hedge) algorithm can be used for the problem of combining on-line algorithms on-line, giving much stronger guarantees than the results of Azar, Y., Broder, A., & Manasse, M. (1993). Proc ACM-SIAM Symposium on Discrete Algorithms (pp. 432–440) when the algorithms being combined occupy a state space of bounded diameter. A generalization of the above, showing how (a simplified version of) Herbster and Warmuth's weight-sharing algorithm can be applied to give a finely competitive bound for the uniform-space Metrical Task System problem. We also give a new, simpler algorithm for tracking experts, which unfortunately does not carry over to the MTS problem.Finally, we present an experimental comparison of how these algorithms perform on a process migration problem, a problem that combines aspects of both the experts-tracking and MTS formalisms.     

Online chasing problems for regular polygons

We consider a server location problem with only one server to move. In this paper we assume that a request is given as a region and that the service can be done anywhere inside the region. Namely, for each request an online algorithm chooses an arbitrary point in the region and moves the server there. Note that if every request is a single point and the server must exactly go there in the given order as conventional server problems, there is no choice for the online player and the problem is trivial. Our main result shows that if the region is a regular n-gon, the competitive ratio of the greedy algorithm is for odd n, and for even n. In particular for a square region, the greedy algorithm turns out to be optimal.     

On Capital Investment

We deal with the problem of making capital investments in machines for manufacturing a product. Opportunities for investment occur over time, every such option consists of a capital cost for a new machine and a resulting productivity gain, i.e., a lower production cost for one unit of product. The goal is that of minimizing the total production costs and capital costs when future demand for the product being produced and investment opportunities are unknown. This can be viewed as a generalization of the ski-rental problem and related to the mortgage problem [3]. If all possible capital investments obey the rule that lower production costs require higher capital investments, then we present an algorithm with constant competitive ratio. If new opportunities may be strictly superior to previous ones (in terms of both capital cost and production cost), then we give an algorithm which is O (min{1+log C , 1+log log P , 1+log M }) competitive, where C is the ratio between the highest and the lowest capital costs, P is the ratio between the highest and the lowest production costs, and M is the number of investment opportunities. We also present a lower bound on the competitive ratio of any on-line algorithm for this case, which is Ω (min{log C , log log P / log log log P , log M / log log M }). This shows that the competitive ratio of our algorithm is tight (up to constant factors) as a function of C , and not far from the best achievable as a function of P and M . Received February 6, 1997; revised November 17, 1997.     

分布式系统中的动态任务分配算法设计

设计一种分布式系统中的动态任务分配算法，并对它所使用的数据结构、实现方法以及稳定性加以讨论。本算法采用双向启动策略，即发送者和接受者都能进行启动、而且能根据系统总负载和任务等待量等自适应地选择启动策略的使用。同时利用阈值和阈长把系统中的节点分为接受节点，负载适中节点和发送节点、采用启发式方法进行任务分配。     

A new measure for the study of on-line algorithms

An accepted measure for the performance of an on-line algorithm is the competitive ratio introduced by Sleator and Tarjan. This measure is well motivated and has led to the development of a mathematical theory for on-line algorithms.We investigate the behavior of this measure with respect to memory needs and benefits of lookahead and find some counterintuitive features. We present lower bounds on the size of memory devoted to recording the past. It is also observed that the competitive ratio reflects no improvement in the performance of an on-line algorithm due to any (finite) amount of lookahead.We offer an alternative measure that exhibits a different and, in some respects, more intuitive behavior. In particular, we demonstrate the use of our new measure by analyzing the tradeoff between the amortized cost of on-line algorithms for the paging problem and the amount of lookahead available to them. We also derive on-line algorithms for theK-server problem on any bounded metric space, which, relative to the new measure, are optimal among all on-line algorithms (up to a factor of 2) and are within a factor of 2K from the optimal off-line performance.     

On the Influence of Lookahead in Competitive Paging Algorithms

We introduce a new model of lookahead for on-line paging algorithms and study several algorithms using this model. A paging algorithm is n-line with strong lookahead l if it sees the present request and a sequence of future requests that contains l pairwise distinct pages. We show that strong lookahead has practical as well as theoretical importance and improves the competitive factors of on-line paging algorithms. This is the first model of lookahead having such properties. In addition to lower bounds we present a number of deterministic and randomized on-line paging algorithms with strong lookahead which are optimal or nearly optimal. Received April 8, 1994; revised May 15, 1995.     

On-Line Competitive Algorithms for Call Admission in Optical Networks

We study the on-line call admission problem in optical networks. We present a general technique that allows us to reduce the problem of call admission and wavelength selection to the call admission problem. We then give randomized algorithms with logarithmic competitive ratios for specific topologies in switchless and reconfigurable optical networks. We conclude by considering full duplex communications. Received August 16, 1997; revised May 11, 1998.     

Competitive randomized algorithms for nonuniform problems

Competitive analysis is concerned with comparing the performance of on-line algorithms with that of optimal off-line algorithms. In some cases randomization can lead to algorithms with improved performance ratios on worst-case sequences. In this paper we present new randomized on-line algorithms for snoopy caching and the spin-block problem. These algorithms achieve competitive ratios approachinge/(e–1) 1.58 against an oblivious adversary. These ratios are optimal and are a surprising improvement over the best possible ratio in the deterministic case, which is 2. We also consider the situation when the request sequences for these problems are generated according to an unknown probability distribution. In this case we show that deterministic algorithms that adapt to the observed request statistics also have competitive factors approachinge/(e–1). Finally, we obtain randomized algorithms for the 2-server problem on a class of isosceles triangles. These algorithms are optimal against an oblivious adversary and have competitive ratios that approache/(e–1). This compares with the ratio of 3/2 that can be achieved on an equilateral triangle.Supported in part by the Center for Discrete Mathematics and Theoretical Computer Science (DIMACS), an NSF Science and Technology Center funded under NSF Contract STC-88-09648 and supported by the New Jersey Commission on Science and Technology.     

On-line scheduling with extendable working time on a small number of machines

Speranza and Tuza [Ann. Oper. Res. 86 (1999) 494-506] studied the on-line problem of scheduling jobs on m identical machines with extendable working time. In this problem, each machine is assumed to have an identical regular working time, which can be extended if necessary. The working time of a machine is the larger one between its regular working time and the total processing time of jobs assigned to it. The objective is to minimize the total working time of machines. They presented an on-line algorithm H_x, with a competitive ratio at most 1.228 for any number of machines by choosing an appropriate parameter x. In this paper we consider a small number of machines. The best choices of x are given for m=2,3,4 and the tight bounds, 7/6, 11/9 and 19/16, respectively, are proved. Among them, the algorithm for m=2 is best possible. We then derive a new algorithm for m=3 with a competitive ratio 7/6.     

On-Line Multi-Threaded Paging

Abstract. In this paper we introduce a generalization of Paging to the case where there are many threads of requests. This models situations in which the requests come from more than one independent source. Hence, apart from deciding how to serve a request, at each stage it is necessary to decide which request to serve among several possibilities. Four different on-line problems arise depending on whether we consider fairness restrictions or not, with finite or infinite input sequences. We study all of them, proving lower and upper bounds for the competitiveness of on-line algorithms. The main competitiveness results presented in this paper state that when no fairness restrictions are imposed it is possible to obtain competitive algorithms for finite and infinite inputs. On the other hand, for the fair case in general there exist no competitive algorithms. In addition, we consider three definitions of competitiveness for infinite inputs. One of them forces algorithms to behave efficiently at every finite stage, while the other two aim at comparing the algorithms' steady-state performances. A priori, the three definitions seem different. We study them and find, however, that they are essentially equivalent. This suggests that the competitiveness results that we obtain reflect the intrinsic difficulty of the problem and are not a consequence of a too strict definition of competitiveness.     

On the power of randomization in on-line algorithms

Against in adaptive adversary, we show that the power of randomization in on-line algorithms is severely limited! We prove the existence of an efficient simulation of randomized on-line algorithms by deterministic ones, which is best possible in general. The proof of the upper bound is existential. We deal with the issue of computing the efficient deterministic algorithm, and show that this is possible in very general cases.A previous version of this paper appeared in the22nd ACM STOC Conference Proceedings. Part of this research was performed while A. Borodin and A. Wigderson were visitors at the International Computer Science Institute, and while G. Tardos was a visitor at the Hebrew University.     

Ordinal algorithms for parallel machine scheduling with nonsimultaneous machine available times

This paper considers ordinal algorithms for parallel machine scheduling with nonsimultaneous machine available times. Two objects of minimizing the latest job completion time and minimizing the latest machine completion time are studied. For the first objective, we present the optimal algorithms for m = 2, 3, 4 machine cases. For m ≥ 5, we propose an algorithm with competitive ratio 2 - 1/(m - 1) while the lower bound is 5/3. For the second objective, the optimal algorithm is also given. Furthermore, for a special case, an algorithm with significantly improved competitive ratio is given.     

New lower bounds for online k-server routing problems

In a k-server routing problem k?1 servers move in a metric space in order to visit specified points or carry objects from sources to destinations. In the online version requests arrive online while the servers are traveling. Two classical objective functions are to minimize the makespan, i.e., the time when the last server has completed its tour (k-Traveling Salesman Problem or k-tsp) and to minimize the sum of completion times (k-Traveling Repairman Problem or k-trp). Both problems, the k-tsp and the k-trp have been studied from a competitive analysis point of view, where the cost of an online algorithm is compared to that of an optimal offline algorithm. However, the gap between the obtained competitive ratios and the corresponding lower bounds have mostly been quite large for k>1, in particular for randomized algorithms against an oblivious adversary.We reduce a number of gaps by providing new lower bounds for randomized algorithms. The most dramatic improvement is in the lower bound for the k-Dial-a-Ride-Problem (the k-trp when objects need to be carried) from to 3 which is currently also the best lower bound for deterministic algorithms.     

Improved Bounds for On-Line Load Balancing

We consider the following load balancing problem. Jobs arrive on-line and must be assigned to one of m machines thereby increasing the load on that machine by a certain weight. Jobs also depart on-line. The goal is to minimize the maximum load on any machine, the load being defined as the sum of the weights of the jobs assigned to the machine divided by the machine capacity. The scheduler also has the option of preempting a job and reassigning it to another machine. Whenever a job is assigned or reassigned to a machine, the on-line algorithm incurs a reassignment cost depending on the job. For arbitrary reassignment costs and identical machines, we present an on-line algorithm with a competitive ratio of 3.5981 against current load , i.e., the maximum load at any time is less than 3.5981 times the lowest achievable load at that time. Our algorithm also incurs a reassignment cost less than 6.8285 times the cost of assigning all the jobs. For arbitrary reassignment costs and related machines we present an algorithm with a competitive ratio of 32 and a reassignment factor of 79.4. We also describe algorithms with better performance guarantees for some special cases of the problem. Received August 24, 1996; revised August 6, 1997.     

Competitive Analysis of the LRFU Paging Algorithm

   Abstract. We present a competitive analysis of the LRFU paging algorithm, a hybrid of the LRU (Least Recently Used) and LFU (Least Frequently Used) paging algorithms. We show that the competitive ratio of LRFU is k +

处理顺序约束的信息物理融合系统静态任务表调度算法

针对异构环境并行计算的静态任务调度问题,以最小化有向无环图 (Directed acyclic graph, DAG)的执行跨度为目标,改变HEFT (Heterogeneous earliest finish time)算法中任务上行权重的计算方法, 获得更加合理的任务顺序排列,提出了一种最早完成时间优先的表调度算法IHEFT (Improvement heterogeneous earliest finish time).该算法在计算任务的上行权重时, 分别计算该任务分配给不同资源的上行权重,取其最小值,比使用所有资源对该任务的平均处理时间进行计算的HEFT算法更为准确. 确定任务的处理顺序后采用最早完成时间越小越优先的策略将任务分配给最优资源,并使得任务的开始执行时间和结束时间满足DAG中有向边的通讯时间约束.通过使用部分文献中的算例数据以及随机生成满足一定结构要求的DAG进行算法测试,将IHEFT与HEFT, CPOP (Critical-path-on-a-processor)和LDCP (Longest dynamic critical path)进行了比较,结果显示IHEFT算法更有效,而且时间复杂度较低.     

LRU Is Better than FIFO

In the paging problem we have to manage a two-level memory system, in which the first level has short access time but can hold only up to k pages, while the second level is very large but slow. We use competitive analysis to study the relative performance of the two best known algorithms for paging, LRU and FIFO. Sleator and Tarjan proved that the competitive ratio of LRU and FIFO is k . In practice, however, LRU is known to perform much better than FIFO. It is believed that the superiority of LRU can be attributed to locality of reference exhibited in request sequences. In order to study this phenomenon, Borodin et al. [2] refined the competitive approach by introducing the concept of access graphs. They conjectured that the competitive ratio of LRU on each access graph is less than or equal to the competitive ratio of FIFO. We prove this conjecture in this paper. Received June 2, 1997; revised January 28, 1998.     

On-Line File Caching

Abstract. Consider the following file caching problem: in response to a sequence of requests for files, where each file has a specified size and retrieval cost , maintain a cache of files of total size at most some specified k so as to minimize the total retrieval cost. Specifically, when a requested file is not in the cache, bring it into the cache and pay the retrieval cost, and remove other files from the cache so that the total size of files remaining in the cache is at most k . This problem generalizes previous paging and caching problems by allowing objects of arbitrary size and cost, both important attributes when caching files for world-wide-web browsers, servers, and proxies. We give a simple deterministic on-line algorithm that generalizes many well-known paging and weighted-caching strategies, including least-recently-used, first-in-first-out, flush-when-full, and the balance algorithm. On any request sequence, the total cost incurred by the algorithm is at most k/(k-h+1) times the minimum possible using a cache of size h ≤ k . For any algorithm satisfying the latter bound, we show it is also the case that for most choices of k , the retrieval cost is either insignificant or at most a constant (independent of k ) times the optimum. This helps explain why competitive ratios of many on-line paging algorithms have been typically observed to be constant in practice.     

Algorithms for the On-Line Travelling Salesman 1

In this paper the problem of efficiently serving a sequence of requests presented in an on-line fashion located at points of a metric space is considered. We call this problem the On-Line Travelling Salesman Problem (OLTSP). It has a variety of relevant applications in logistics and robotics. We consider two versions of the problem. In the first one the server is not required to return to the departure point after all presented requests have been served. For this problem we derive a lower bound on the competitive ratio of 2 on the real line. Besides, a 2.5 -competitive algorithm for a wide class of metric spaces, and a 7/3 -competitive algorithm for the real line are provided. For the other version of the problem, in which returning to the departure point is required, we present an optimal 2 -competitive algorithm for the above-mentioned general class of metric spaces. If in this case the metric space is the real line we present a 1.75 -competitive algorithm that compares with a \approx 1.64 lower bound. Received November 12, 1997; revised June 8, 1998.