This thesis discusses an analogue method of function generation using a magnetic drum. The function generator is suitable for use in a pulse- position-modulation analogue computer which operates on a time-sequential principle and in real time. The functions are approximated by linear and parabolic arcs, and the break points are represented by pulse posit ions. These pulse positions are stored on a magnetic drum. The function generator converts these pulse positions into the original function by using precise sweep amplifiers and gating circuits. The number of functions that are available is limited only by the size of the drum. The method of function generation proposed here allows for very simple programming of the computer. Accuracies approaching one per cent were obtained in tests using very inadequate sweep amplifiers. Accuracies of the order of one-tenth of one per cent are believed to be possible by using the proposed function generator that is described at the end of this thesis. This thesis also describes the synchronization and counting circuits that are required to control the operation of the computer. Also the circuits required to write the information pulses on the drum are described. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

We investigate the properties of an unsupervised neural network which uses simple Hebbian learning and negative feedback of activation in order to self-organise. The negative feedback circumvents the well-known difficulty of positive feedback in Hebbian learning systems which causes the networks' weights to increase without bound. We show, both analytically and experimentally, that not only do the weights of networks with this architecture converge, they do so to values which give the networks important information processing properties: linear versions of the model are shown to perform a Principal Component Analysis of the input data while a non-linear version is shown to be capable of Exploratory Projection Pursuit. While there is no claim that the networks described herein represent the complexity found in biological networks, we believe that the networks investigated are not incompatible with known neurobiology. However, the main thrust of the thesis is a mathematical analysis of the emergent properties of the network; such analysis is backed by empirical evidence at all times.

Thesis (Ph. D.)--Oregon State University, 1982. / Typescript (photocopy). Includes bibliographical references (leaves 300-311). Also available on the World Wide Web.