D.E. Goldberg, K. Deb, and D. Theirens. Toward a better understanding of mixing in genetic algorithms. In Belew and Booker [21], pages 190--195.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....two individuals do not differ (Gi = Gopt) the fitness (cost) of Gi is f = 0. In this work we only want to use a minimization problem. Because this test problem is similar to the standard one max problem it is easy to solve for mutation based GEAs, but somewhat harder for recombination based GAs [13]. 4 Performance of Evolution Strategies and Adjustment of Parameters In this section, after a short introduction into the functionality of evolution strategies, we present an investigation into the adjustment of ES parameters for the one max tree problem when using the NetKey encoding. The ....

D. E. Goldberg, K. Deb, and D. Thierens. Toward a better understanding of mixing in genetic algorithms. Journal of the Society of Instument and Control Engineers,

....block di#cult, and 6. ensuring that building blocks exchange to form better solutions. These items emphasize the importance of GAs design choices, including initialization of the population, parameters like population size, crossover and mutation probabilities, selection strategy etc. In [32], it is expressed that although there have been a substantial progress in the first 5 items almost no work has considered the building block exchange. While it is critical to supply, grow and select the building blocks it is also very crucial to combine a building block in one structure with the ....

D.E. Goldberg, K. Deb, and D. Thierens. Toward a better understanding of mixing in genetic algorithms. Technical Report IlliGAL Report No. 92009.

....pressure caused by migration. This leads to accurate predictions of the number of generations until convergence. In addition, understanding the e ect of the migration policies on the selection pressure is important because excessively slow or fast convergence rates may cause the search to fail (Goldberg et al. 1993; Thierens and Goldberg, 1993) If selection is too weak the population may drift aimlessly for a long time, and the quality of the solutions found is not likely to be good. On the other hand, rapid convergence is desirable, but an excessively fast convergence may cause the EA to converge ....

....leads to accurate predictions of the number of generations until convergence. In addition, understanding the e ect of the migration policies on the selection pressure is important because excessively slow or fast convergence rates may cause the search to fail (Goldberg et al. 1993; Thierens and Goldberg, 1993). If selection is too weak the population may drift aimlessly for a long time, and the quality of the solutions found is not likely to be good. On the other hand, rapid convergence is desirable, but an excessively fast convergence may cause the EA to converge prematurely to a suboptimal solution. ....

Goldberg, D. E., K. Deb, and D. Thierens: 1993, `Toward a better understanding of mixing in genetic algorithms'. Journal of the Society of Instrument and Control Engineers 32(1), 10-16.

....over succeeding generations. The convergence rate of genetic algorithm is largely determined by the selection pressure. Higher selection pressures result higher convergence rates. Genetic algorithms are able to identify optimal or near optimal solutions under a wide range of selection pressure [9]. However, if the selection rate is too low, the convergence rate will be slow, and the genetic algorithm will unnecessarily take longer to find the optimal solution. If the selection pressure is too high, there is an increased change of genetic algorithm prematurely converging to an incorrect ....

Goldberg, D. E., Deb., Thierens, D. (1993). Toward better understanding of mixing in genetic algorithms. Journal of the Society of Instrument and Control Engineers, 32(1) 10-16

....single point crossover, etc. Proportionate selection has various drawbacks, the scaling problem being the foremost. Single point crossover is known to disrupt good building blocks and thereby increase the convergence time as well as population size required. It has been shown in recent works (Goldberg, Thierens, Deb, 1993; Thierens, 1994; Thierens Goldberg, 1993) that ensuring e ective building block (BB) mixing is an integral part of ecient GA design. These studies also showed that this could be achieved through a tight linkage of the set of alleles belonging to a BB. Based upon this concept many novel ....

....various drawbacks, the scaling problem being the foremost. Single point crossover is known to disrupt good building blocks and thereby increase the convergence time as well as population size required. It has been shown in recent works (Goldberg, Thierens, Deb, 1993; Thierens, 1994; Thierens Goldberg, 1993) that ensuring e ective building block (BB) mixing is an integral part of ecient GA design. These studies also showed that this could be achieved through a tight linkage of the set of alleles belonging to a BB. Based upon this concept many novel competent GA designs have been proposed which can be ....

Goldberg, D., Thierens, D., & Deb, K. (1993). Toward a Better Understanding of Mixing in Genetic Algorithms. J. Soc. Instrument and Control Engineers, 32 (1), 10-16.

....simpli cations in either the algorithm or the problem model. Therefore, the practical value of the current theoretical results on parameter settings is unclear. 4 There are some theoretical investigations on the optimal population size [50] 132] 60] 52] or optimal operator probabilities [54], 131] 10] 108] however, these results were based on simple function optimization problems and their applicability for other types of problems is limited. A general drawback of the parameter tuning approach, regardless of how the parameters are tuned, is based on the observation that a run ....

D.E. Goldberg, K. Deb, and D. Theirens. Toward a better understanding of mixing in genetic algorithms. In Belew and Booker [21], pages 190-195.

....pressure, but high selective pressures also lead to faster convergence and therefore leave less time for the mixing process to perform its task. Cross competition arises when building blocks from different partitions compete with each other as opposed to the competition within the partitions [GDT93], where a partitioning divides the search space in a set of disjunct regions. In case of a binary search space such a partition can be described by means of a set of non overlapping schemata f fff f , where is the wild card symbol and f represents a fixed allele (either 0 or 1) Two types of ....

D.E. Goldberg, K. Deb, and D. Thierens. Towards a better understanding of mixing in genetic algorithms. Journal of the Society for Instrumentation and Control Engineers, SICE, 32(1):10--16, 1993.

.... Mathematical and General, 224] Journal of Structural Engineering ASCE, 850, 910] Journal of the Institute of Systems, Control, and Information Engineers (Japan) 205, 380, 996] Journal of the Operational Research Society, 352] Journal of the Society of Instrument and Control Engineers, [375, 509, 571] KI Lexikon, 190] Kikai Gijutsu Kenkyusho Shoho, 997] Machine Learning, 313, 386, 391, 516, 1065] Machine Learning Journal, 216] Mech. Syst. Signal Process. UK) 193] Methods of Information in Medicine, 881] Microprocessing and Microprogramming, 1073] Neural Computing and ....

....S. 198] Das, Rajarshi, 1065] Dasgupta, Dipankar, 199, 200, 201, 704, 705, 706, 707, 708, 709] Dastidar, D. Ghosh, 202] David, E. 963] Authors 17 Davidge, Robert, 203] Davidor, Yuval, 204, 205] Davis, Lawrence, 206, 207, 208, 209, 210] Davis, Thomas Elder, 973] Deb, Kalyanmoy, [217, 375, 376, 378, 381] Deboeck, Guido, 218, 219] deFigueiredo, Rui J. P. 944] Delaney, B. 239] Denham, M. J. 797] Deodhar, D. 538] Deugo, Dwight, 220, 221] Dhawan, Atam P. 222, 223, 657, 806] Dike, B. A. 834, 837] Dissanayake, M. W. M. G. 1072] Diver, D. A. 224] Dix, T. I. 815] Dobnikar, ....

[Article contains additional citation context not shown here]

David E. Goldberg, Kalyanmoy Deb, and Dirk Thierens. Toward a better understanding of mixing in genetic algorithms. Journal of the Society of Instrument and Control Engineers, 32(1):10--16, 1993. also as [?] ga:Goldberg93a.

....in practical situations. This middle ground approach has also been quite productive. For example, it has lead to formulating population sizing equations which are compact and reasonably easy to understand (Goldberg et al. 1991) and to produce recipes on how to properly set the parameters of a GA (Goldberg et al. 1993). Very recently Stephens and Waelbroeck (Stephens and Waelbroeck 1997, Stephens and Waelbroeck 1999) have produced a new schema theorem which, unlike previous results which concentrated on schema survival and disruption, makes the effects and the mechanisms of schema creation explicit. This ....

Goldberg, D. E., K. Deb and D. Thierens (1993). Toward a better understanding of mixing in genetic algorithms. Journal of the Society of Control Engineers (SICE) 32(1), 10--16.

....these highlyfit building blocks to form larger, highly fit building blocks, and repeat this process recursively until a complete solution is found. Numerous studies have focused on defining building blocks (Goldberg, Korb, and Deb 1989; Haynes 1997) encouraging effective mixing of building blocks (Thierens and Goldberg 1993), and understanding building block dynamics in a GA (Forrest and Mitchell 1992; Spears 1998; van Kemenade 1997; Wu and Lindsay 1996; Wu, Lindsay, and Riolo 1997) In order for this strategy to work, a fine balance must be maintained between the exploitation of known building blocks of information ....

Thierens, D. and D. E. Goldberg (1993). Toward a better understanding of mixing in genetic algorithms. Journal Society of Control Engineers 32(1), 10--16.

....pressure caused by migration. This leads to accurate predictions of the number of generations until convergence. In addition, understanding the effect of the migration policies on the selection pressure is important because excessively slow or fast convergence rates may cause the search to fail (Goldberg, Deb, Thierens, 1993; Thierens Goldberg, 1993) If selection is too weak the population may drift aimlessly for a long time, and the quality of the solutions found is not likely to be good. On the other hand, rapid convergence is desirable, but an excessively fast convergence may cause the EA to converge ....

....accurate predictions of the number of generations until convergence. In addition, understanding the effect of the migration policies on the selection pressure is important because excessively slow or fast convergence rates may cause the search to fail (Goldberg, Deb, Thierens, 1993; Thierens Goldberg, 1993). If selection is too weak the population may drift aimlessly for a long time, and the quality of the solutions found is not likely to be good. On the other hand, rapid convergence is desirable, but an excessively fast convergence may cause the EA to converge prematurely to a suboptimal solution. ....

Goldberg, D. E., Deb, K., & Thierens, D. (1993). Toward a better understanding of mixing in genetic algorithms. Journal of the Society of Instrument and Control Engineers, 32 (1), 10--16.

....in this manner by now (see [15] for a summary of the results) 3. 3 Dimensional Analysis The observation that the exact Markov model is isomorphic to the associated EA but offers only little chances to extract important aspects has led to the idea of approaching EAs via dimensional analysis [21,22]. This methodology is borrowed from engineering sciences [23] Dimensional analysis tries to identify the important dimensions or key features of a complex system and establishes a functional relationship between them. When applied to evolutionary algorithms, isolated measures for iterated ....

D. E. Goldberg, K. Deb, and D. Thierens. Toward a better understanding of mixing in genetic algorithms. Journal of the Society for Instrumentation and Control Engineers (SICE), 32(1):10--16, 1993.

....interesting optimizers to find solutions to such problems. The search for the optimum usually involves two steps: growing the building blocks in the population and mixing these building blocks in order to obtain the global optimal solution. The mixing of building blocks can be difficult for a GA [GDT93, TG93]. One approach to increase the capabilities of GA s to handle such mixing difficulties is to measure correlations between bit values. Two methods that take this approach are the GEMGA [Kar96] and the building block filtering [vK96] Both methods estimate fitness contributions of bits. The GEMGA ....

D.E. Goldberg, K. Deb, and D. Thierens. Towards a better understanding of mixing in genetic algorithms. Journal of the Society for Instrumentation and Control Engineers, SICE, 32(1):10--16, 1993.

No context found.

Goldberg D.E., Deb K., & Thierens D. (1993). Toward a better understanding of mixing in genetic algorithms. Journal of the Society for Instrumentation and Control Engineers, SICE Vol.32, No.1 pp.10-16.

....ECGA to a non azeotropic binary working uid power cycle optimization problem. The optimal power cycle obtained improved the cycle eciency by 2.5 over that existing cycles, thus illustrating the capabilities of ECGA in solving real world problems. 1 Introduction It has been shown in recent works [1, 2, 3] that ensuring e ective building block (BB) mixing is an integral part of ecient GA design. These studies also showed that this could be achieved through a tight linkage of the set of alleles belonging to a BB. Based upon this concept many novel competent GA designs have been proposed which can be ....

D. Goldberg, D. Thierens, and K. Deb, \Toward a Better Understanding of Mixing in Genetic Algorithms, " J. Soc. Instrument and Control Engineers, vol. 32, no. 1, pp. 10-16, 1993.

....on GA s performance. These isolated studies are worthwhile and have provided useful guidelines for choosing GA parameters, such as population size (Goldberg, Deb, and Clark, 1992; Harik et al. 1997) and control maps for operator probabilities (Goldberg, Deb, and Thierens, 1992; Thierens and Goldberg, 1993). In order to observe the interactions of various GA parameters, empirical studies have also been used (De Jong, 1975; Eshelman and Schaffer, 1993; Schaffer et al. 1989; Wu, Lindsay, and Riolo, 1997) To study the dynamics of these interactions, more sophisticated stochastic models using Markov ....

....However, we ignore such special implementations in this study. function is difficult for GAs to solve, GAs must be allowed more number of points to search from. Although there exists no clear study specifying what would cause GA difficulty, the following few factors have been suggested elsewhere (Goldberg, 1993; Horn, Goldberg, and Deb, 1994) 1. Multi modality 2. Deception 3. Isolation 4. Collateral noise Multi modality causes difficulty to any search and optimization method, because of the presence of a number of false attractors. For some algorithms (such as gradient descent methods) only a few ....

Goldberg, D. E., Deb, K., and Thierens, D. (1993). Toward a better understanding of mixing in genetic algorithms. Journal of SICE, 32(1), 10--16. Proceedings of the Fourth International Conference on Genetic Algorithms. (pp. 190--195).

No context found.

D.E. Goldberg, K. Deb and D. Thierens, "Toward a better understanding of mixing in genetic algorithms", Journal of the Society for Instrumentation and Control Engineers, 32, 1, 10--16, 1993.

....algorithms, and O(2 l ) for proportionate selection) and the convergence times for the uniformly salient problems like the OneMax problem (resp. O( p l) and (O(l ln l) The modeling approach taken here is to build simple facetwise models and then interrelate them in a dimensionless way (Goldberg, Deb, Thierens, 1993). An important extension is to include the e ect of building block disruption: both the BinInt and the OneMax allows us to study the e ects of the tness or salience distribution on 6 the convergence speed, while bracketing the in uence of building block survival and disruption. Incorporating this ....

Goldberg D.E., Deb K., & Thierens D. (1993). Toward a better understanding of mixing in genetic algorithms.

....on the scalability properties of the simple genetic algorithm. In section 5, we consider whether some straightforward extension to the simple GA might improve the scaling problems. Finally, we discuss the ramifications of our findings. Part of the work reported here was introduced in Thierens and Goldberg (1993) and Thierens (1995) c #1999 by the Massachusetts Institute of Technology Evolutionary Computation 7(4) 331 352 D. Thierens 2 Inductive Bias of Genetic Algorithms Genetic algorithms are search procedures, and, by definition, this implies that they make some assumptions about the underlying ....

....normal runtime of the GA. A different approach to linkage learning is taken by the messy GA and its successors, which first identify the building blocks and, subsequently, tightly code them. Research on the messy GA is continuing and has already achieved interesting results (Goldberg et al. 1989; Goldberg et al. 1993; Kargupta, 1996) 3 Selection versus Mixing Holland (1975) identified building blocks as the fundamental units of GA processing primarily by examining the schema theorem. A somewhat generalized version of the theorem may be written: m#h; t 1##m#h; t###h; t##1 , ##h; t##,with##h; t# the ....

[Article contains additional citation context not shown here]

Goldberg, D. E., Deb, K. and Thierens, D. (1993). Toward a better understanding of mixing in genetic algorithms. Journal of the Society for Instrumentation and Control Engineers, 32(1):10--16.

....on the scalability properties of the simple genetic algorithm. In section 5 we consider whether some straightforward extension to the simple GA might improve the scaling problems. Finally we discuss the rami cations of our ndings. Part of the work reported here has been published in (Thierens Goldberg, 1993; Thierens, 1995) 1 2 Inductive Bias of Genetic Algorithms Genetic algorithms are search procedures and by de nition this implies that they have to make some assumptions about the underlying structure of the search space which guides their decision making in order to be more ecient than ....

....the GA will reliably converge to the global optimum. In the next paragraph we will check experimentally the dimensional relation. 4.1. 2 Empirical veri cation All experiments in this section and the next are performed with fully deceptive trap functions with signal value equal to 1 3 (Deb Goldberg, 1993). For a single parameter combination 50 runs are tried and called successful when at least 49 of them converge to the global optimum. We use uniform crossover with allele wise swapping probability px = 0:5, no mutation and block selection (i.e. the n s best strings in the current population get ....

[Article contains additional citation context not shown here]

Goldberg D.E., Deb K., & Thierens D. (1993). Toward a better understanding of mixing in genetic algorithms. Journal of the Society for Instrumentation and Control Engineers, SICE Vol.32, No.1 pp.10-16.

No context found.

D. E. Goldberg, K. Deb, and D. Thierens. Toward a better understanding of mixing in genetic algorithms. Journal of the Society of Instrument and Control Engineers, 32#1#:10#16, 1993.

....algorithms, and O(2 l ) for proportionate selection) and the convergence times for the uniformly salient problems like the OneMax problem (resp. O( p l) and (O(l ln l) The modeling approach taken here is to build simple facetwise models and then interrelate them in a dimensionless way (Goldberg, Deb, Thierens, 1993). An important extension is to include the effect of building block disruption: both the BinInt and the OneMax allows us to study the effects of the fitness or salience distribution on 6 the convergence speed, while bracketing the influence of building block survival and disruption. Incorporating ....

Goldberg D.E., Deb K., & Thierens D. (1993). Toward a better understanding of mixing in genetic algorithms.

....reviews that decomposition and explores the type of rough analytics required to piece the puzzle together. 3. 1 A GA design decomposition A decomposition of the problem of designing a selectorecombinative genetic algorithm that reflects the current state of affairs has been presented elsewhere (Goldberg, 1993; Goldberg, Deb, Clark, 1992) 1. Know what you re processing: building blocks (BBs) 2. Ensure there is an adequate initial supply of BBs; 3. Ensure that necessary BBs are expected to grow; 4. Ensure that BB decisions are well made; 5. Solve problem of bounded BB difficulty; 6. Ensure that BBs ....

....algorithm that solve hard problems quickly. Recently, simple models of each of the portions of the GA design decomposition were assembled to predict the region of effective convergence for a simple GA operating on an easy problem in terms of the selection pressure s and crossover probability p c (Goldberg, Deb, Thierens, 1993). Figure 1 shows the shape of the theoretically predicted control map, which shows how to set up a GA over a large range of GA parameters. Moreover, in these experiments good results were obtained in a subquadratic number of function evaluations: since the number of generations is O(log n) n the ....

[Article contains additional citation context not shown here]

Goldberg, D. E., Deb, K., & Thierens, D. (1993). Toward a better understanding of mixing in genetic algorithms.

No context found.

D.E. Goldberg, K. Deb, and D. Theirens. Toward a better understanding of mixing in genetic algorithms. In Belew and Booker [21], pages 190--195.

No context found.

Goldberg, D., Deb, K., and Thierens, D., "Toward A Better Understanding of Mixing in Genetic Algorithms," Journal of the Society of Control Engineers, Vol. 32, No. 1, January 1993, pp. 10-16.

First 50 documents

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