Đang chuẩn bị liên kết để tải về tài liệu:
Báo cáo khoa học: "Language Learning in Massively-Parallel Networks"

In vision, it is sometimes possible to design networks from a task analysis of ~he problem, aided by the homogeneity of the domain. For example, Sejnowski & Hinton (1986) designed a network that can separate figure from ground for shapes with incomplete bounding contours. Constructing a network is much more difficult in an inhomogeneous domain like natural language. This problem has been partially overcome by the discovery of powerful learning algorithms that allow the strengths of connection in a network to be shaped by experience; . | FORUM ON CONNECTIONISM Language Learning in Massively-Parallel Networks Terrence J. Sejnowski Biophysics Department Johns Hopkins University Baltimore MD 21218 PANELIST STATEMENT Massively-parallel connectionist networks have traditionally been applied to constraint-satisfaction in early visual processing Ballard Hinton Sej nowski 1983 but are now being applied to problems ranging from the TravelingSalesman Problem to language acquisition Rumelhart McClelland 1986 . In these networks knowledge is represented by the distributed pattern of activity in a large number of relatively simple neuron-like processing units and computation is performed in parallel by the use of connections between the units. A network model can be programmed by specifying the strengths of the connections or weights on all the links between the processing units. In vision it is sometimes possible to design networks from a task analysis of the problem aided by the homogeneity of the domain. For example Sejnowski Hinton 1986 designed a network that can separate figure from ground for shapes with incomplete bounding contours. Constructing a network is much more difficult in an inhomogeneous domain like natural language. This problem has been partially overcome by the discovery of powerful learning algorithms that allow the strengths of connection in a network to be shaped by experience that is a good set of weights can be found to solve a problem given only examples of typical inputs and the desired outputs Sejnowski Kienker Hinton 1986 Rumelhart Hinton Williams 1986 . Network learning will be demonstrated for the problem of converting unrestricted English text to phonemes. NETtalk is a network of 309 processing units connected by 18 629 weights Sejnowski Rosenberg 1986 . It was trained on the 1 000 most common words in English taken from the Brown corpus and achieved 98 accuracy. The same network was then tested for generalization on a 20 000 word dictionary without further training it was 80 .