Chen, H., Flann, N., Watson, D.: Parallel genetic simulated annealing: A massively parallel SIMD algorithm. IEEE Transactions on Parallel and Distributed Systems 9 (1998) 126--136  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... nish. In this study, we decided to use the technique of GAs because it has already been extensively used for solving scheduling problems such as Job Shop [12, 43, 18, 15, 46] the Travelling Salesman problem [45, 25] instruction scheduling [8, 7] or scheduling on multiprocessor en4 vironments [30, 48, 16, 33, 11, 17, 53] and has proven to be successful. GA performances have been compared with other optimisation techniques such as Tabu Search [6] or Simulated Annealing [51] and the results show that GAs are never bad on average for a large variety of problems : GA is said to be a robust approach [24] For ....

.... GA performances have been compared with other optimisation techniques such as Tabu Search [6] or Simulated Annealing [51] and the results show that GAs are never bad on average for a large variety of problems : GA is said to be a robust approach [24] For instance, previous studies [27] quoted in [11]) have shown that GAs are far less sensitive to parameter values than Simulated Annealing whose performances are largely dependant on the ne tuning of the parameters. Furthermore, because there is a single solution that is modi ed over time, Simulated Annealing is in essence a serial algorithm ....

H. Chen, N.S. Flann, and D. Watson. Parallel genetic simulated annealing : A massively parallel SIMD algorithm. IEEE Transactions on Parallel and Distributed Systems, 9(2):126136, February 1998.

....channel of each Transputer. Speculative trees usually have a huge communication overhead and systolic algorithm works with few processors and reduce the SA efficacy. Some algorithms combine SA with genetic algorithms and neuronal networks and use the intrinsic parallelism of these methods [2, 12]. However, they are not pure parallel SA algorithms. Recently, we have proposed MPSA [5] which is a methodology that leads to massive parallelism of the SA algorithm. Any parallel SA algorithm derived through MPSA has an executing time that in the best case takes only the last temperature cycle ....

Chen, H., Flann, N.S. and Watson, D.W., 1998. Parallel genetic simulated annealing: a massively parallel SIMD algorithm. IEEE Trans on Parallel and Distributed Systems, Vol. 9, No. 2, pp. 126 -- 136.

....Min#min or GA. Because SA allows poorer solutions to be accepted at intermediate stages, it allows some very poor solutions in the initial stages, from which it can never recover (see Section 4) GSA: The Genetic Simulated Annealing (GSA) heuristic is a combination of the GA and SA techniques [7, 36]. In general, GSA follows procedures similar to the GA outlined above. However, for the selection process, GSA uses the SA cooling schedule and system temperature and a simplified SA decision process for accepting or rejecting a new chromosome. Specifically, the initial system temperature was set ....

....and reduced to 900 of its current value for each iteration. Whenever a mutation or crossover occurs, the new chromosome is compared with the corresponding original chromosome. If the new makespan is less than the original makespan plus the system temperature, then the new chromosome is accepted [7, 36]. Otherwise, the original chromosome survives to the next iteration. Therefore, as the system temperature decreases, it is again more difficult for poorer solutions to be accepted. The two stopping criteria used were either (a) no change in the elite chromosome in 150 iterations or (b) 1000 total ....

H. Chen, N. S. Flann, and D. W. Watson, Parallel genetic simulated annealing: A massively parallel SIMD approach, IEEE Trans. Parallel Distrib. Comput. 9, 2 (Feb. 1998), 126#136.

....generational genetic algorithm (genGA) the steadystate genetic algorithm (ssGA) 13] and the cellular genetic algorithm (cGA) 8] Second, the paper reports the improvementachieved on already known results 2 for similar problem instances. We compare the results of our experiments to those of [2] and [6] The outline of the paper is as follows: Section 2 presents a short overview of the basic working principles of genetic algorithms. Section 3 presents the maximum cut problem, the error correcting code design problem, and the minimum tardy task problem. The problem s encoding, the tness ....

.... problems represent a broad spectrum of the challenging intractable problems in the areas of graph theory [9] coding theory [7] and scheduling [4] All three problems were chosen because of their practical use and the existence of some preliminary work in applying genetic algorithms to solve them [2], 3] and [6] 3 The experiments for graph and scheduling problems are performed with different instances. The rst problem instance is of moderate size, but nevertheless, isachallenging exercise for any heuristic. While the typical problem size for the rst instance is about twenty, the ....

[Article contains additional citation context not shown here]

H. Chen, N. S. Flann, and D. W. Watson. Parallel genetic simulated annealing: a massively parallel SIMD algorithm. #### ############ ## ######## ### ########### #######, pages 805-811, vol. 9, number 2, February 1998.

....the generational genetic algorithm (genGA) the steadystate genetic algorithm (ssGA) 13] and the cellular genetic algorithm (cGA) 8] Second, the paper reports the improvement achieved on already known results for similar problem instances. We compare the results of our experiments to those of [2] and [6] The outline of the paper is as follows: Section 2 presents a short overview of the basic working principles of genetic algorithms. Section 3 presents the maximum cut problem, the error correcting code design problem, and the minimum tardy task problem. The problem s encoding, the tness ....

.... problems represent a broad spectrum of the challenging intractable problems in the areas of graph theory [9] coding theory [7] and scheduling [4] All three problems were chosen because of their practical use and the existence of some preliminary work in applying genetic algorithms to solve them [2], 3] and [6] The experiments for graph and scheduling problems are performed with different instances. The rst problem instance is of moderate size, but nevertheless, is a challenging exercise for any heuristic. While the typical problem size for the rst instance is about twenty, the ....

[Article contains additional citation context not shown here]

H. Chen, N. S. Flann, and D. W. Watson. Parallel genetic simulated annealing: a massively parallel SIMD algorithm. IEEE transactions on parallel and distributed systems, pages 805-811, vol. 9, number 2, February 1998.

.... experimental results, obtained from a sequential implementation of the approach that simulates the behaviour of the cellular automaton, show signi cant better performances with respect to C4.5 and comparable performances with respect to CGP [4] The method follows other recent hybrid methods [8, 2], that incorporate simulated annealing into genetic algorithms, and the Cellular Genetic Algorithm proposed in [22] Our approach, however, is the rst proposal that couples cellular genetic programming and simulated annealing for classifying databases. A cellular automaton is composed of a set ....

....nal value when MaxNumberOfGeneration steps have been executed. select(t i ; t 0 ; t 1 ; temperature) rst chooses between t 0 and t 1 the one having the best tness. Suppose it is t 0 . Then, t i is replaced by t 0 only if fitness(t 0 ) fitness(t i ) temperature. This deterministic criterion [2] has been shown to be less expensive and to perform equivalently to the random technique. 6 Implementation and Experimental Results In this section we present the experiments and results obtained by a preliminary implementation of the method on a sequential machine. The CGP SA classi er has been ....

H.Chen, N.S.Flann and D.W.Watson (1998). Parallel Genetic Simulated Annealing: A Massively Parallel SIMD Algorithm. In IEEE Transaction on Parallel and Distributed Systems, vol. 9, No.2.

....[103, 205, 61, 29] Europhysics Letters, 237, 241] Guangxue Xuebao, 245] Helvetica Physica Acta, 141] IEE Proceedings C: Generation, Transmission and Distribution, 76, 187] IEE Proceedings Comput. Digital Tech. 223] IEEE Expert, 20] IEEE Trans. Parallel Distrib. Syst. USA) [125] IEEE Transaction on Information Technology in Biomedicine, 232] IEEE Transactions on Computer Aided Design of Integrated Circuits and Systems, 261, 264] IEEE Transactions on Magn. 102] IEEE Transactions on Magnetics, 201, 248] IEEE Transactions on Power Systems, 191, 98, 210, 217, ....

....E. 238, 239] Buydens, Lutgarde M. C. 100, 218] Cai, Ziyong, 205] Caiti, A, 102] Candia, Alfredo, 227] Cao, Y. J. 97, 53] Carter, Bob, 10, 183, 66] Celli, G. 190] Chakraborty, M. 108] Chakraborty, U. K. 108] Chan, Shu Park, 38] Chen, B. H. 128] Chen, C. C. 123] Chen, H. [155, 125] Chen, Hsinchun, 127] Chen, W. C. 123] Chen, Wen chin, 47] Chen, Yen Wei, 133] Cheng, K. S. 128] Cheu, Wen Chin, 77] Chiaberge, M. 92] Chikhani, A. Y. 181] Cho, Hyeon Joong, 222, 63] Choi, Doo Hyun, 222, 63] Choudhary, A. 13] Clark, David E. 24] Cohoon, James P. 179, ....

[Article contains additional citation context not shown here]

H. Chen, N. S. Flann, and D. W. Watson. Parallel genetic simulated annealing: a massively parallel SIMD algorithm. IEEE Trans. Parallel Distrib. Syst. (USA), 9(2):126-136, 1998. yCCA34045/98 ga98aHChen.

....of SA. The heuristic stops when there is no change in the makespan for 150 iterations or the system temperature reaches zero. Most tests ended with no change in the makespan for 150 iterations. GSA: The Genetic Simulated Annealing (GSA) heuristic is a combination of the GA and SA techniques [4, 23]. In general, GSA follows procedures similar to the GA outlined above. GSA operates on a population of 200 chromosomes, uses a Min min seed in four out of eight initial populations, and performs similar mutation and crossover operations. However, for the selection process, GSA uses the SA cooling ....

H. Chen, N. S. Flann, and D. W. Watson, "Parallel genetic simulated annealing: A massively parallel SIMD approach," IEEE Transactions on Parallel and Distributed Computing, Vol. 9, No. 2, Feb. 1998, pp. 126--136.

No context found.

Chen, H., Flann, N., Watson, D.: Parallel genetic simulated annealing: A massively parallel SIMD algorithm. IEEE Transactions on Parallel and Distributed Systems 9 (1998) 126--136

No context found.

H. Chen, N. S. Flann, and D. W. Watson, "Parallel genetic simulated annealing: A massively parallel SIMD algorithm," IEEE Trans. on Parallel and Distributed Systems, vol. 9, no. 2, pp. 126--136, 1998.

No context found.

H. Chen, N.S. Flann, and D. Watson. Parallel genetic simulated annealing : A massively parallel SIMD algorithm. IEEE Transactions on Parallel and Distributed Systems, 9(2):126136, February 1998.

No context found.

H. Chen, N.S. Flann, and D.W. Watson. Parallel genetic simulated annealing: A massively parallel SIMD algorithm. IEEE Transactions on Parallel and Distributed Systems, 9(2):126--136, 1998.

No context found.

H. Chen, N. S. Flann, and D. W. Watson, "Parallel genetic simulated annealing: A massively parallel SIMD algorithm," IEEE Trans. Parallel and Distributed System, vol. 9, no. 2, pp. 126--136, Feb. 1998.

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