R. L. Boehning R. M. Butler and B. E. Gillet, A Parallel Integer Linear Programming Algorithm, European Journal of Operations Research 34 (1988).  Home/Search   Document Not in Database   Summary   Related Articles   Check

....programs (LPs) although this is the most typical scenario. A third layer, called the BiCePS Linear Integer Solver (BLIS) implements LP based branch and bound algorithms. A number of techniques for developing scalable, parallel branch and bound algorithms have been proposed in the literature [6, 11, 12, 16, 17, 32]. However, we know of no previous work speci cally addressing the development of scalable algorithms for data intensive applications. Standard techniques for parallel branch and bound break down when applied to BCP, primarily because they all depend on the ability to easily shuttle search tree ....

R. L. Boehning R. M. Butler and B. E. Gillet, A Parallel Integer Linear Programming Algorithm, European Journal of Operations Research 34 (1988).

....data structures based on a unique di erencing scheme which is memory ecient. This scheme for storing the tree allows us to handle problems much larger than we otherwise could. A number of techniques for developing scalable parallel branch and bound algorithms have been proposed in the literature [3, 9, 10, 15, 17, 39]. However, we know of no previous work speci cally addressing the development of scalable algorithms for data intensive applications. Standard techniques for parallel branch and bound break down when applied to BCP, primarily because they all depend on the ability to easily shuttle search tree ....

R.L. Boehning R.M. Butler and B.E. Gillet, A Parallel Integer Linear Programming Algorithm, European Journal of Operations Research 34 (1988).

....integer linear programs was crudely constructed by the authors without the additional optimizations such as sparse systems, etc. Furthermore, much of our computational process is naturally parallelizable and should benefit immensely from techniques such as parallel simplex [20] and parallel ILP [1, 3]. After running the generators to find the IB MAP, we turned to actually approximating the probabilities. In [45] the runtime of the cost sharing scheme compared favorably with stochastic simulation (as implemented in CaBeN [11] and our runtime was better than the stochastic simulation for 100 ....

Rochelle L. Boehning, Ralph M. Butler, and Billy E. Gillett. A parallel integer linear programming algorithm. European Journal of Operational Research, 34:393--398, 1988.

....the solution of mixed integer programs of sizes and with speeds that did not seem possible ten years ago. There are many reasons to believe that parallel computing may provide the opportunity for even greater advances. Related work in parallel distributed mixed integer program solution appears in [1, 2, 3, 4, 5, 6]. Parino (Parallel Mixed Integer Optimizer) is a parallel software system that we have developed for solving mixed integer programs (MIPs) Parino is an evolving system, College of Computing, Georgia Institute of Technology y Industrial and Systems Engineering, Georgia Institute of Technology ....

R.L. Boehning, R.M. Butler, B.E. Gillett, "A Parallel Integer Linear Programming Algorithm," European J. Operations Research 34, 393-398, 1988.

....type of parallel machine in order to ensure a good trade off between communication and computation, resulting in a good performance for the parallel algorithm. The inherent parallelism in B B method is suitable to implementation either on MIMD shared memory multiprocessor machines 1 see e.g. [15, 16, 17, 18, 19] or on MIMD distributed memory coarse grained multiprocessors 2 (see the paper of Dehne, Ferreira and Rau Chaplin [20] for more references) Two major problems arise in this approach: i) Load balancing: how to manage the sizes of the problems assigned to individual processors; the site may vary ....

R.L.Boehning, R. Butler, and B. Gillett, "A parallel integer linear programming algorithm," European Journal of Operational Research, vol. 34, pp. 393--398, 1988.

.... programming (implementation of a B B on the one hand, choosing a parallelization version on the other hand, has led several research scientists to suggest programming environments dedicated to B B [23, 63, 59, 18] or to develop generic B B programs of which they have outlined the basic aspects [46, 18, 20, 31, 3]. We present here a similar approach. We designed a specialized library for B B algorithms. The idea is, of course, to facilitate writing B B for any application and to make it possible : to test several parallelization parameters (data structures, load balancing strategies, different ....

R. Boehning, R. Butler, and B. Gillett. A parallel integer linear programming algorithm. (34):393--398, 1988.

....integer linear programs was crudely constructed by the authors without the additional optimizations such as sparse systems, etc. Furthermore, much of our computational process is naturally parallelizable and should benefit immensely from techniques such as parallel simplex [13] and parallel ILP [1, 2]. 5 RELATED WORK The work on term computation [8] and related papers are extremely relevant to this paper. The skewness assumption made there, or a weaker version of it, also make our method applicable. In a sense, these methods complement each other, and it should be interesting to see whether ....

Rochelle L. Boehning, Ralph M. Butler, and Billy E. Gillett. A parallel integer linear programming algorithm. European Journal of Operational Research, 34:393--398, 1988.

....the most crudest of implementations. There are still by far many more optimizations to be made whether they are optimizations from the Bayesian networks or just general implementational optimizations such as from Operations Research tools. Parallel computation techniques can also be considered [2, 1]. And most interestingly, we may begin looking at domain dependent factors involved with what the Bayesian network itself is attempting to model. Acknowledgements I would like to thank Jak Kirman and Ann Nicholson for providing me with the necessary data from the Mobile Target Location problem. ....

Rochelle L. Boehning, Ralph M. Butler, and Billy E. Gillett. A parallel integer linear programming algorithm. European Journal of Operational Research, 34:393--398, 1988.

....integer linear programs was crudely constructed by the authors without the additional optimizations such as sparse systems, etc. Furthermore, much of our computational process is naturally parallelizable and should benefit immensely from techniques such as parallel simplex [19] and parallel ILP [2, 3]. We only did a few experiments to compare the runtime of our algorithms to exact algorithms. Clearly, in some cases exact algorithms would be faster, e.g. on poly trees (unless we also modify our algorithms to take advantage of this special case, which is possible) Nevertheless, we ran across ....

Rochelle L. Boehning, Ralph M. Butler, and Billy E. Gillett. A parallel integer linear programming algorithm. European Journal of Operational Research, 34:393--398, 1988.

....were single processes on a workstation. There is nothing in this architecture to prevent the major tasks from actually consisting of a group of parallel processes executing on multiprocessor systems. In fact, existing systems for genetic algorithms such as genesis [16] integer linear programming [4, 1] and A [17] searches can be naturally decomposed and distributed. 3.1.1 Genetic Algorithms As we mentioned earlier, the class of solutions based on genetic algorithms [15, 23] is a nondeterministic approach to reasoning with Bayesian networks. Genetic algorithms take a small sample from the ....

Rochelle L. Boehning, Ralph M. Butler, and Billy E. Gillett. A parallel integer linear programming algorithm. European Journal of Operational Research, 34:393--398, 1988.

No context found.

Boehning, R.L., Butler, R.M., and Gillet, B.E.: A Parallel Integer Linear Programming Algorithm. European Journal of Operations Research 34, 393, 1988

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC