| G.Z. Sun, H.H. Chen, C.L. Giles, Y.C. Lee, and D. Chen. Connectionist pushdown automata that learn context-free grammars. In Proceedings of the International Joint Conference on Neural Networks, Washington D.C., 1990. |
....modeling for spoken dialog systems remains unfortunately widely unsolved. It was claimed for a long time that connectionist architectures, which are efficient for static pattern classification, were not able to deal with symbol sequences. Now, experimental results as well as theoretical studies [5] have demonstrated the ability of recurrent networks [6] to process word sequences. Actually, two main problems seem to prevent still the use of connectionist models: their time and data consuming nature during the learning phase, their relative difficulties to model complex linguistic ....
....must consider the linguistic context (word order constraints) to understand the HVH sentences. A recurrent network with 12 hidden neurons succeeded completely on this second experiment. Once again, this result is not surprising: recurrent networks can parse word sequences with a rigid syntax [5]. The most interesting result is provided by the last experiment: it shows that recurrent nets can associate various linguistic constraints to understand flexible word order sentences. Two nets, with respectively 15 and 16 hidden neurons, present indeed an interesting accuracy and selectivity ....
Sun G.Z. et al (1990), Connectionist pushdown automata that learn context-free grammars, Proc. IJCNN'90, Washigton DC, USA, pp. 577-580.
....a global view of the parameterization problem which complements the local view of gradient descent methods. 1. Introduction A number of researchers have studied the induction of context free grammars by neural networks. Many have used an external stack and negative evidence ( Giles et al. 1990) (Sun et al. 1990), Das et al. 1992) Das et al. 1993) Mozer and Das, 1993) Zheng et al. 1994) Some have used more standard architectures and only positive evidence ( Wiles and Elman, 1995) Rodriguez et al. ta) In all cases, only very simple context free languages have been learned. 2 Table 1: ....
Sun, G. Z., Chen, H. H., Giles, C. L., Lee, Y. C., and Chen, D. (1990). Connectionist pushdown automata that learn context-free grammars. In Caudill, M., editor, Proceedings of the International Joint Conference on Neural Networks, pages 577--580. Lawrence Earlbaum, Hillsdale, NJ.
No context found.
G.Z. Sun, H.H. Chen, C.L. Giles, Y.C. Lee, D. Chen, Connectionist Pushdown Automata that Learn Context-Free Grammars, Proceedings of the International Joint Conference on Neural, Washington D.C., Lawrence Erlbaum Pub., Vol I, p. 577 (1990).
.... to have powerful capabilities for modeling many computational structures; an excellent discussion of recurrent neural network models and references can be found in (Hertz, Krogh Palmer 1991) To learn grammars, we use a second order recurrent neural network (Lee et al. 1986; Giles et al. 1990; Sun et al. 1990; Pollack 1991) The network architecture is illustrated in figure 1. This net has N recurrent hidden neurons labeled S j ; L special, Neural Network Rule Extraction 6 nonrecurrent input neurons labeled I k ; and N 2 Theta L real valued weights labeled W ijk . As long as the number of input ....
Sun, G.Z., Chen, H.H., Giles, C.L., Lee, Y.C. & Chen, D. (1990). Connectionist Pushdown Automata that Learn Context-Free Grammars. Proceedings of the International Joint Conference on Neural Networks (pp. 577-580). Hillsdale: Lawrence Erlbaum.
No context found.
G.Z. Sun, H.H. Chen, C.L. Giles, Y.C. Lee, and D. Chen. Connectionist pushdown automata that learn context-free grammars. In Proceedings of the International Joint Conference on Neural Networks, Washington D.C., 1990.
No context found.
G. Z. Sun, H. H. Chen, C. L. Giles, Y. C. Lee, and D. Chen. Connectionist pushdown automata that learn context-free grammars. In M. Caudill, editor, Proceedings of the International Joint Conference on Neural Networks, pages 577--580. Lawrence Earlbaum, Hillsdale, NJ, 1990.
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