transcription, транскрипция: [ ˈnjuərəl ]
adjective (Science and Technology) In computing jargon: modelled on the arrangement of neurons in the brain and nervous system; used especially in neural network (or neural net), a computer system which is designed to simulate the human brain in its ability to 'learn' probabilistically and carry out complex processes simultaneously at a number of different nodes. Etymology: A figurative use of the adjective neural. History and Usage: The development of computer neural networks was founded on the work of mathematicians studying neurophysiology as a model for the construction of automata from the late forties onwards; it was not until the eighties, though, that computer scientists announced that they had succeeded in building a computer which worked on the neural principle. The basic principle underlying the neural net computer is that of connectivity; essentially this means doing away with a central processor in favour of a number of simple calculating elements which work in parallel and are connected in patterns similar to those of human neurons and synapses. Such a system, unlike the digital computer, can solve problems even when there are minor inaccuracies in the starting data, and can also be 'trained' to use a technique for reaching correct solutions based on trial and error. The neural net computer is therefore seen as one of the most promising areas of AI research in the early nineties. A number of special neural networks will be designed and interlinked to create a neural computer...Research into neural computing is now a multi-million pound scientific endeavour. The Times 25 Mar. 1989, p. 5 We're also looking at advanced neural nets and doing quite a lot of work on VLSI (Very Large Scale Integration), to make sure that the memory we develop is properly structured and packaged in a chip. CU Amiga Apr. 1990, p. 91 There's something big just below the surface of neural-net technology, something real big. PC Magazine June 1990, p. 170