H. Bersini and B. Renders, 1994, "Hybridizing genetic algorithms with hill--climbing methods for global optimization: Two possible ways", IEEE International Symposium Evolutionary Computation, pp. 312--317, Orlando, Fl.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the common TSP formulation which considers all the edges in a tour. This setting was settled down after some experiments. 4 A Hybrid Genetic Algorithm A genetic algorithm hybridized with local optimizations is called a hybrid GA. A great many studies about hybridization of GAs were proposed [49][32]. Sammon s Mapping Since the microarray data are virtually real valued vectors in a high dimensional space, we map them into the two dimensional space in order to use the 2D natural crossover, which operates on chromosomes encoded by 2D graphic images. We chose the Sammon s mapping described ....

J. M. Renders and H. Bersini. Hybridizing genetic algorithms with hill-climbing methods for global optimization: Two possible ways. In Proceedings of the First IEEE Conference on Evolutionary Computation, pages 312-317, 1994.

....space with n taxa has n elements if all possibilities are considered. Our GA provides an alternative search method to nd a good order of taxa addition. A genetic algorithm hybridized with local optimizations is called a hybrid GA. A considerable number of studies about hybridization of GAs [30] [29] [19] have been proposed. Figure 3 shows a typical steady state hybrid genetic algorithm. In the next subsection, we describe each part of the hybrid GA that we used for this work. 4 Yong Hyuk Kim et al. fastDNAml( n : the nal number of taxa i : the number of taxa in the current tree ....

J. M. Renders and H. Bersini. Hybridizing genetic algorithms with hill-climbing methods for global optimization: two possible ways. In Proceedings of the First IEEE Conference on Evolutionary Computation, pages 312-317, 1994.

.... the PGI research) Genetic algorithms may also fail to find optimal solutions either due to ruggedness of the fitness landscape [11]or because the problem is deceptive [8] Methods for addressing these problems include modified crossover operators, hybrid search (say, with hill climbing methods [16]) and random changes in the representation of the search space [10] Unfortunately, many of the features for which these methods are designed apply to the attribute search problem (e.g. premature convergence, high ruggedness) Unfortunately, again, none of these methods presents a silver bullet ....

J.-M. Renders and H. Bersini. Hybridizing genetic algorithms with hillclimbing methods for global optimization: Two possible ways. In IEEE World Congress on Computational Intelligence, volume 1, pages 312-- 317, 1994.

....global nature of their search) but rather inaccurate and inefficient in finding the global minimum, several modifications have been proposed to exploit their advantage and compensate for their shortcoming. Of special interest is the work by Renders and Bersini, who proposed two type of hybrid GAs (Renders and Bersini, 1994). The fist type consists in interwoving GAs with Hill Climbing techniques (GA HC) the solution selection no longer depends on the instantaneous evaluation of the fitness function applied to the solution but rather applied to a refinement of the solution obtained via Hill Climbing techniques. The ....

J.M. Renders and H. Bersini. Hybridizing Genetic Algorithms with Hill Climbing Methods for Global Optimization: Two Possible Ways. In First IEEE Conference on Evolutionary Computation, pages 312--317, 1994.

....meet some practical constraints or the original model was obtained from a noisy environment. Although such directed optimisation techniques still serve as the major optimisation tool in control systems engineering, following an a priori direction may not lead to globally optimised model parameters (Renders and Bersini, 1994), as the unknown multidimensional error surface is usually multimodal. Further, it is difficult to identify an optimal order or structure if two or more separate parameters (such as repeated poles) have a similar effect on the system output. To achieve good reduction tractability and quality, ....

....tractability and robustness in global optimisations by slightly trading off precision in a nondeterministic polynomial (NP) manner. Thus, compared with the non NP exhaustive search or dynamic programming techniques, EAs offer an exponentially reduced search time. Compared with Powell s technique, Renders and Bersini (1994) have shown that EAs offer double accuracy and reliability in multidimensional optimisations. The evolutionary methods, in their various forms, have successfully been applied to problems related to model reduction, such as system identification and linearisation (Kristinsson and Dumont, 1992; ....

[Article contains additional citation context not shown here]

Renders, J. M. and Bersini, H. 1994. `Hybridizing genetic algorithms with hill-climbing methods for global optimisation: two possible ways', Proc 1st IEEE Int. Conf. Evolutionary Computation, First IEEE World Cong. Computational Intelligence, Orlando, 1, 312-317.

....and the manufacturing cell formation problem. The hybrid GA is shown to be a very efficient and effective optimization method, especially when a one to one mapping is enforced (i.e. Larmarckian evolution is employed) II. HYBRIDIZING GAS WITH LOCAL IMPROVEMENT PROCEDURES Many researchers [1,3,6,12,14,18] have shown that GAs perform well for global searching 3 because they are capable of quickly finding and exploiting promising regions of the search space, but they take a relatively long time to converge to a local optimum. Local improvement procedures, e.g. two opt switching for combinatorial ....

....region of the search space, but are typically poor global searchers. Local improvement procedures have been incorporated into GAs in order to improve the algorithm s performance through what could be termed learning. Such hybrid GAs have been used successfully to solve a wide variety of problems [1,6,12,14]. There are two basic models of evolution that can be used to incorporate learning into a GA: the Baldwin Effect and Lamarckian evolution. The Baldwin Effect allows an individual s fitness (phenotype) to be determined based on learning, i.e. the application of local improvement. Like natural ....

[Article contains additional citation context not shown here]

H. Bersini and B. Renders. Hybridizing genetic algorithms with hill-climbing methods for global optimization: Two possible ways. In 1994 IEEE International Symposium Evolutionary Computation, pages 312--317, Orlando, Fl, 1994.

....arithmetical crossover into multi parent operator; parents x 1 ; x 2 ; x r may produce a family of offspring: y = a 1 x 1 a 2 x 2 : a r x r , for any choice of (a i ) 2 [0; 1] r such that a 1 a 2 : a r = 1. The simplex crossover operator proposed by Renders and Bersini (1994) for numerical optimization problems can also be viewed as another generalization of arithmetical crossover; this crossover operator involves computing the centroid of group of parents and moving from the worst individual beyond the centroid point. More precisely, the operator selects k 2 ....

.... x i = k Gamma 1) and computes the reflected point y (i.e. the offspring) obtained from the worst one: y = c ( c Gamma w) This operator includes also a few extra options, which are based on the outcome of comparisons between the values of f( y) f( w) and f( b) for details, see Renders and Bersini, 1994). Some of these ideas were embodied earlier in the evolutionary procedure called scatter search (Glover, 1977) The process generates initial populations by screening good solutions produced by heuristics. The points used as parents are then joined by linear combinations with contextdependent ....

Renders, J.-M. and H. Bersini (1994). Hybridizing genetic algorithms with hill-climbing methods for global optimization: Two possible ways. In Z. Michalewicz, J. D. Schaffer, H.-P.

....allows one to preset the maximum number of (not necessarily unique) solutions that are evaluated. In this study, we preset the maximum number of generations and then determine at which generation the GA finds the best solution. 3 Incorporation of a Local Improvement Procedure Many researchers [7, 8, 16, 17, 30, 34, 35] have shown that GAs perform well for global searching because they are capable of quickly finding and exploiting promising regions of the search space, but they take a relatively long time to converge to a local optimum. For example, Figure 1 shows three replications of a typical GA run for the ....

J-M. Renders and H. Bersini. Hybridizing genetic algorithms with hill-climbing methods for global optimization: Two possible ways. In The First IEEE Conference on Evolutionary Computation, 1994.

....with respect to being trapped in local minima (due to the global nature of their search) but rather inaccurate and inefficient in finding the global minimum, several modifications have been proposed to exploit their advantage and compensate for their shortcoming. Of special interest is the work by Renders and Bersini (1994), who proposed two type of hybrid GAs. The fist type consists in interwoving GAs with Hill Climbing techniques (GA HC) the solution selection no longer depends on the instantaneous evaluation of the fitness function applied to the solution but rather applied to a refinement of the solution ....

Renders, J.M. and Bersini, H. (1994). Hybridizing Genetic Algorithms with Hill Climbing Methods for Global Optimization: Two Possible Ways. In First IEEE Conference on Evolutionary Computation, pages 312--317.

....a local improvement operator which, as shown in [Houck et al. 1995b] can greatly enhance the performance of the genetic algorithm. Many researchers have shown that GAs perform well for a global search but perform very poorly in a localized search [Davis 1991; Michalewicz 1994; Houck et al. 1995a; Bersini and Renders 1994]. GAs are capable of quickly finding promising regions of the search space but may take a relatively long time to reach the optimal solution. Both the float genetic algorithm (FGA) and binary genetic algorithm (BGA) were run 10 times with different random seeds. The simulated annealing (SA) ....

Bersini, H. and Renders, B. 1994. Hybridizing genetic algorithms with hill-climbing methods for global optimization: Two possible ways. In 1994 IEEE International Symposium Evolutionary Computation, Orlando, Fl, pp. 312--317.

....than local optimization techniques. It can only recombine good guesses hoping that one recombination will have a better fitness than both of its parents 3 . Because of this limitation, many researchers have combined GAs with other optimization techniques to develop hybrid genetic algorithms [35, 14, 15, 36, 37, 38, 17, 21, 20, 39]. The purpose of such hybrid systems is to speed up the rate of convergence while retaining the ability to avoid being easily entrapped at a local optimum. Although local optimization in a hybrid often results in a faster convergence, it has been shown that too much local optimization can ....

....with probability 1, to lower ranking chromosomes with probability 0. Examples of elite based hybrid GA include G bit improvement on GA and our simplex GA hybrid. The partition based hybrid GA was introduced in Renders Bersini s work, even though they did not give this architecture a name [35]. As we shall see later in Section VIII, the choice between these two architectures can have a significant impact on the performance of a hybrid GA system. B. Simplex GA Hybrid Approaches In this section, we describe two approaches for incorporating the simplex method into the GA as an additional ....

[Article contains additional citation context not shown here]

J. Renders and H. Bersini, "Hybridizing genetic algorithms with hill-climbing methods for global optimization: Two possible ways," In Proceedings of the First IEEE Conference on Evolutionary Computation, pp. 312--317, Orlando, FL, June 1994.

....than local optimization techniques. It can only recombine good guesses hoping that one recombination will have a better fitness than both of its parents 2 . Because of this limitation, many researchers have combined GAs with other optimization techniques to develop hybrid genetic algorithms [30, 14, 15, 31, 32, 33, 17, 20]. The purpose of such hybrid systems is to speed up the rate of convergence while retaining the ability to avoid being easily entrapped at a local optimum. Although local optimization in a hybrid often results in a faster convergence, it has been shown that too much local optimization can ....

....the pure real coded GA, 2) a 45 simplex GA hybrid, 3) a 100 concurrent probabilistic simplex, 4) the R B hybrid. The parameters of R B hybrid are those reported in their paper for a function maximization problem: 0.5 simplex probability, 0.2 crossover probability, and 0. 2 average probability [30]. We chose the percentage of simplex reproductions in our hybrid approach such that the portion of simplex reproduction in the total reproduction is about the same as that of R B s approach: 45 hybrid: P Theta w Gamma N ) p c Theta P (1 Gamma w) P Theta w Gamma N ) 147 Theta 0:45) ....

[Article contains additional citation context not shown here]

J. Renders and H. Bersini, "Hybridizing genetic algorithms with hill-climbing methods for global optimization: Two possible ways," In Proceedings of the First IEEE Conference on Evolutionary Computation, pp. 312--317, Orlando, Florida, June 1994.

....p m . P4.5. n : n 2. air96.tex; 14 03 1996; 19:57; no v. p.40 REAL CODED GAs 41 P4.6. If n N Then Go to P4.2. P5. If t is not a multiple of u Then P5.1. Pop(t 1) Pop(t) P5.2. Go to P2. P6. For i = 1; N Do e i (t) 0. t : t 1. Go to P2. Another attempt was the GA Simplex (Renders et al. 1994). The basic idea was to increase the local tuning of the RCGA by using a crossover operator called simplex crossover, which simulates the behaviour of a hill climbing method called Simplex. 5. Tools for the Analysis of the RCGA Some authors attempted to generalize the schema theorem for the real ....

Renders, J. M. & Bersini, H. (1994). Hybridizing Genetic Algorithms with HillClimbing Methods for Global Optimization: Two Possible Ways. Proc. of The First IEEE Conference on Evolutionary Computation, 312-317.

....3 stage evolution (LIMS) In Section V, empirical results are given. Finally, concluding remarks are made in Section VI. II. A REVIEW OF SPX This section gives a brief review of SPX [Tsutsui 99] Higuchi 00] Here, note that there is another crossover which is also named Simplex Crossover [Renders 94] But it is completely different scheme from our idea. The SPX operator uses multi parent, n 1 parental vectors X , i = 0, 1, n for recombination. These (n 1) vectors form a simplex in R n . Then this simplex is expanded in each direction (X i O) to some extent, where O is the center of ....

Renders, J. and Bersini, H.: Hybridizing genetic algorithms with hill-climbing methods for global optimization:two possible ways, Proc. of IEEE Int. Conf. on Evolutionary Computation, pp. 312-317 (1994).

No context found.

H. Bersini and B. Renders, 1994, "Hybridizing genetic algorithms with hill--climbing methods for global optimization: Two possible ways", IEEE International Symposium Evolutionary Computation, pp. 312--317, Orlando, Fl.

No context found.

Renders, J-M. and Bersini, H.: Hybridizing genetic algorithms with hill-climbing methods for global optimization: two possible ways, Proc. of the First IEEE Conference on Evolutionary Computation, pp. 312-317 (1994).

No context found.

Renders, J-M. and Bersini, H.: Hybridizing genetic algorithms with hill-climbing methods for global optimization: two possible ways, Proc. of the First IEEE Conference on Evolutionary Computation, pp. 312-317 (1994).

No context found.

Renders, J. M. and Bersini, H. Hybridizing genetic algorithms with hill-climbing methods for global optimisation: two possible ways, Proc.

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