S. Oliker, M. Furst, and O. Maimon. A distributed genetic algorithm for neural network design and training. Complex Systems, 6(5):459--477, 1992.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....and natural representation. The key problem (other than being trapped in a local minimum) with BP and other traditional training algorithms is the choice of a correct architecture (number of hidden nodes and connections) This problem has been tackled by the evolutionary approach in many studies [4, 14, 17, 21, 24, 39, 40, 41]. In some of these studies, weights and architectures are evolved simultaneously. The major disadvantage to the EANN approach is it is computationally expensive, as the evolutionary approach is normally slow. To overcome the slow convergence of the evolutionary approach to ANN, hybrid techniques ....

S. Oliker, M. Furst, and O. Maimon. A distributed genetic algorithm for neural network design and training. Complex Systems, 6(5):459--477, 1992.

....and what search operators should be used in evolving architectures in Section III D. A. The Direct Encoding Scheme Two different approaches have been taken in the direct encoding scheme. The first separates the evolution of architectures from that of connection weights [24] 150] 153] 154] [165], 167] 169] 170] The second approach evolves architectures and connection weights simultaneously [149] 179] 180] 182] 185] 200] This section will focus on the first approach. The second approach will be discussed in Section III D. In the first approach, each connection of an ....

....simultaneously [149] 179] 180] 182] 185] 200] This section will focus on the first approach. The second approach will be discussed in Section III D. In the first approach, each connection of an architecture is directly specified by its binary representation [24] 150] 153] 154] [165], 167] 169] 170] 202] For example, an matrix can represent an ANN architecture with nodes, where indicates presence or absence of the connection from node to node . We can use to indicate a connection and to indicate no connection. In fact, can represent real valued connection weights from ....

[Article contains additional citation context not shown here]

S. Oliker, M. Furst, and O. Maimon, "A distributed genetic algorithm for neural network design and training," Complex Syst., vol. 6, no. 5, pp. 459--477, 1992.

....Bull. Sci. Assoc. Ing. Electr. Inst. Electrotech. Montefiore, 633] Bulletin of the Polish Academy of Sciences Chemistry, 100] Cancer Letters, 313] Chemometrics and Intelligent Laboratory Systems, 149, 495, 802] Chinese Journal of Advanced Software Research, 421] Complex Systems, [65, 786, 843, 845, 852, 874, 916] Comput. Ind. Eng. 387] Comput. Struct. UK) 382] Computer, 402] Computer Applications in the Biosciences (CABIOS) 109, 224] Computer Design, 220] Computers and Electronics in Agriculture, 515, 534] Computers Industrial Engineering, 255, 655] Computers Operations Research, ....

....[240] Fujii, T. 498] Fujimoto, Yoshiji, 561] Fujita, S. 36] Fukuda, Toshio, 147, 249, 257, 294, 310, 552, 569, 700, 701, 702, 703, 704, 705, 706, 707, 708] Fukumi, M. 336, 444, 466] Fukumi, Minoru, 709, 710] Fullmer, Brad, 838] Funabiki, N. 499] Funabiki, S. 498] Furst, M. [852, 853] 18 Genetic algorithms and neural networks Furuhashi, Takeshi, 500] Furuya, Tatsumi, 84, 342, 371, 396, 986] Gaborski, Roger S. 856] Gabriele, Gabriele, 91] Galbiati, R. 321] Gali c, Elvis, 337] Gallagher, John C. 611, 612] Gallagher, N. B. 149] Gant, V. 332] Gao, Xinbo, ....

[Article contains additional citation context not shown here]

S. Oliker, M. Furst, and O. Maimon. A distributed genetic algorithm for neural network design and training. Complex Systems, 6(5):459--477, 1992. y(CCA 42891/93) ga:Oliker92a. Bibliography 83

....to every journal article included in this bibliography. The list is arranged in alphabetical order by the name of the journal. Annual Report of the German Society for Aviation and Space ight, 171] APL Quote Quad, 422] Biological Cybernetics, 173] BioSystems, 39, 533] Complex Systems, [148, 462] Complex Systems (USA) 270] Comput. Econ. 156] Computational mechanics, 320] Computer Aided Design, 523] Computer Applications in the Biosciences (CABIOS) 199, 295] Computer Graphics, 492] Computers in Chemical Engineering, 244] Computers Chemistry, 231, 347] Computers ....

....Flann, N. S. 94] Fleming, Peter J. 131] Fogarty, Terence C. 133, 223, 398, 547, 399, 548, 400] Fogel, David B. 401] Fraga, E. S. 244] Freisleben, Bernd, 402] Fujikawa, Hideji, 235] Fukuda, Toshio, 304] Fukuyama, Y. 147, 227] Fukuyama, Yoshikazu, 134, 224, 245, 281] Furst, M. [462, 463] Gammack, John G. 547, 548] Garcia, F. 534] Garis, Hugo de, 14, 10, 17, 18, 19, 25, 50, 51] Gates, Jr. George H. 91, 111] G.Attolico, 343] Gaylord, Richard J. 27] Geary, R. A. 392] Genco, A. 203] Gers, F. 35] Geyer Schulz, Andreas, 422] Ghosh, R. K. 274] Ghoshray, S. ....

[Article contains additional citation context not shown here]

S. Oliker, M. Furst, and O. Maimon. A distributed genetic algorithm for neural network design and training. Complex Systems, 6(5):459-477, 1992. yCCA 42891/93 ga:Oliker92a. 76 Distributed genetic algorithms

.... 682, 683, 690, 691] APL Quote Quad, 392] Applied Artificial Intelligence, 258, 587] Archiv fur Elektrotechnik, 79] Atoms, Molecules and Clusters, 418] BioEngineering, 223] Biological Cybernetics, 57, 488] Biopolymers, 447] CC AI, 199] Clinical Chemistry, 308] Complex Systems, [139, 222, 245, 247, 248, 326, 509, 516, 532, 540, 595, 597] Composites Engineering, 116] Computer Physics Communications, 571] Computers Industrial Engineering, 524, 546] Computers Mathematics with Applications, 353] Design Theory and Methodology, 562] Daedalus, 323] Electronics Letters, 37, 530, 556] Electronics Weekly, 144] Eng. ....

....P. 120] Frankhauser, Pierre, 219] Frazer, J. H. 345] Freeman, L. M. 375] Freeman, Ray, 565] Frenzel, James F. 220, 221] Freund, Harald, 222] Freyer, Stephan, 223] Frieder, Ophir, 635] Fukuda, Toshio, 224, 225, 226, 227, 228, 229, 230, 231] Fullmer, Brad, 496] Furst, M. [532] Furuhashi, Takeshi, 674] Furuya, Tatsumi, 719] Gall, A. Le, 182] Gallagher, John C. 85] Galletly, J. E. 239] Gammack, John G. 194] Gandham, Ravi V. 685] Garai, I. 240] Garcia, F. 615] Garigliano, Roberto, 526, 16] Garis, Hugo de, 716, 717, 718, 719, 720] Gemmill, D. ....

[Article contains additional citation context not shown here]

S. Oliker, M. Furst, and O. Maimon. A distributed genetic algorithm for neural network design and training. Complex Systems, 6(5):459--477, 1992. y(CCA 42891/93) ga:Oliker92a.

....parents from the population based on their fitness. 5. Apply search operators to the parents and generate offspring which form the next generation. Figure 6: A typical cycle of the evolution of architectures. Considerable research on evolving ANN architectures has been carried out in recent years [33, 42, 45, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 149, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 138, 213, 214, 215, 216, 118, 130, 127, 217, 218, 219, 220, 221, 222, 223, 128, 224, 225]. Most of the research has concentrated on the evolution of ANN topological structures. Relatively little has been done on the evolution of node transfer functions, let al..one the simultaneous evolution of both topological structures and node transfer functions. In this paper, we will analyze the ....

....architectures is beneficial and what search operators should be used in evolving architectures in Section 3.4. 3.1 The Direct Encoding Scheme Two different approaches have been taken in the direct encoding scheme. The first separates the evolution of architectures from that of connection weights [154, 153, 150, 24, 170, 169, 165, 167]. The second approach evolves architectures and connection weights simultaneously [179, 180, 182, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 149, 198, 199, 200] This section will focus on the first approach. The second approach will be discussed in Section 3.4. In the first ....

[Article contains additional citation context not shown here]

S. Oliker, M. Furst, and O. Maimon, "A distributed genetic algorithm for neural network design and training," Complex Systems, vol. 6, no. 5, pp. 459--477, 1992.

.... s starting weights and genetically searches, instead, for an appropriate network topology. Most methods that use GAs to optimize a network topology use backpropagation to train each network s weights. Of these methods, many directly encode each connection in the network (Miller et al. 1989; Oliker et al. 1992; Schiffmann et al. 1992) These methods are relatively straightforward to implement, and are good at fine tuning small networks (Miller et al. 1989) however, they do not scale well since they require very large matrices to represent large networks (Yao, 1993) Other techniques (Harp et al. ....

....encoding schemes can evolve different sets of parameters along with the network s topology and have been shown to have good scalability (Yao, 1993) Regent differs from both the direct and indirect methods in that it does not explicitly encode its networks. Some techniques (Koza Rice, 1991; Oliker et al. 1992) evolve both the architecture and connection weights at the same time; however, the combination of the two levels of evolution greatly increases the search space. Regent differs mainly from both GA and non GA (Fahlman Lebiere, 1989; Mezard Nadal, 1989; Frean, 1990) network growing algorithms ....

Oliker, S., Furst, M., & Maimon, O. (1992). A distributed genetic algorithm for neural network design and training. Complex Systems, 6:459--477.

....the next generation. Figure 6: A typical cycle of the evolution of architectures. Reprinted with permission from Ref. 1. X. Yao: Evolutionary Artificial Neural Networks 19 Because of advantages of the evolutionary design of architectures, a lot of research has been carried out in recent years [37, 46, 49, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100]. However, almost all the research only deals with the topological structure of EANNs and little has been done on the evolution of node transfer functions let al..one the evolution of both topological structures and node transfer functions. We will concentrate on the evolution of topological ....

....interchangeably with the term topological structure in these two sections. Section 3.3 discusses the evolution of node transfer functions briefly. 3. 1 The Direct Encoding Scheme In the direct encoding scheme, each connection in an architecture is directly specified by its binary representation [82, 81, 78, 29, 98, 97, 93, 95]. For example, an N Theta N matrix C = c ij ) N ThetaN can represent an architecture with N nodes, where c ij indicates presence or absence of the connection from node i to node j. We can use c ij = 1 to indicate a connection and c ij = 0 no connection. In fact, c ij can even be connection ....

[Article contains additional citation context not shown here]

S. Oliker, M. Furst, and O. Maimon. A distributed genetic algorithm for neural network design and training. Complex Systems, 6:459--477, 1992.

....for appropriate network topologies. Most methods that use genetic algorithms to optimize a network topology are similar to Regent in that they also use backpropagation to train each network s weights. Of these methods, many directly encode each link in the network (Miller, Todd, Hegde, 1989; Oliker, Furst, Maimon, 1992; Schiffmann, Joost, Werner, 1992) These methods are relatively straightforward to implement, and are good at fine tuning small networks (Miller et al. 1989) however, they do not scale well since they require very large matrices to represent all the links in large networks (Yao, 1993) Other ....

....the network, such as the number of hidden layers, the number of hidden nodes at each layer, etc. These indirect encoding schemes can evolve different sets of parameters along with the network s topology and have been shown to have good scalability (Yao, 1993) Some techniques (Koza Rice, 1991; Oliker et al. 1992) evolve both the architecture and connection weights at the same time; however, the combination of the two levels of evolution greatly increases the search space. Regent mainly differs from genetic algorithm based training methods in that it is designed for knowledge based neural networks. Thus ....

Oliker, S., Furst, M., & Maimon, O. (1992). A distributed genetic algorithm for neural network design and training. Complex Systems, 6, 459--477.

....is not directly optimized by the evolutionary algorithm but rather a result of the cascade algorithm. In [4, 9, 64, 68, 70] the weights and topology of recurrent neural networks are determined, Zhang [82] optimizes Sigma Pi networks A quite unusual approach has been proposed by Oliker et al. [55] where the search for the optimal neural network is done separately for every single neuron, i.e. separately in different genetic algorithms working together to finally build an optimal network structure. The underlying idea is that the search space is drastically reduced in comparison to genetic ....

S. Oliker and M. Furst. A distributed genetic algorithm for neural network design and training. Complex systems, 6:459--477, 1992.

....the children generated above, and obtain the next generation. Figure 3: A typical cycle of the evolution of architectures. X. Yao: Evolutionary Artificial Neural Networks 13 Because of advantages of the evolutionary design of architectures, a lot of research has been carried out in recent years [36, 45, 48, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96]. However, almost all the research only deals with the topological structure of EANNs, little has been done on the evolution of node transfer functions, let al..one the evolution of both topological structures and node transfer functions. We will concentrate on the evolution of topological ....

....interchangeably with the term topological structure in these two sections. Section 3.3 discusses the evolution of node transfer functions briefly. 3. 1 The Direct Encoding Scheme In the direct encoding scheme, each connection in an architecture is directly specified by its binary representation [78, 77, 74, 27, 94, 93, 89, 91]. For example, an N Theta N matrix C = c ij ) N ThetaN can represent an architecture with N nodes, where c ij indicates presence or absence of the connection from node i to node j. We can use c ij = 1 to indicate a connection and c ij = 0 no connection. In fact, c ij can even be connection ....

[Article contains additional citation context not shown here]

S. Oliker, M. Furst, and O. Maimon. A distributed genetic algorithm for neural network design and training. Complex Systems, 6:459--477, 1992.

No context found.

S. Oliker and M. Furst, "A distributed genetic algorithm for neural network design and training," Complex Systems, vol. 6, no. 5, pp. 459--477, 1992.

No context found.

S. Oliker, M. Furst, and O. Maimon. A Distributed Genetic Algorithm for Neural Network Design and Training. Complex Systems, 6(5):459--477, Oct 1992.

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