MIT Press.

Research report RC 294, IBM, Yorktown Heights, New York

In Yovitts, M. and Cameron, S. (eds.) Self-organizing Systems. New York: Pergamon

This paper briefly reviews the evidence for multistore theories of memory and points out some difficulties with the approach. An alternative framework for human memory research is then outlined in terms of depth or levels of processing. Some current data and arguments are reexamined in the light of this alternative framework and implications for further research considered.

Group Report 54-15, Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, Mass

Teddington Conference

"MANY EFFORTS have been made to discriminate,  categorize, and quantitate patterns, and  to reduce them into a usable machine language.  The results have ordinarily been methods or devices  with a high degree of specificity. For example, some  devices require a special type font; others can read  only one type font; still others require magnetic ink. We have an interest in decision-making circuits  with the following qualities: (1) measurable high reliability  in decision making, (2) either a high or a low  reliability input, and (3) possibly low reliability components.  The high specificity of the devices and  methods mentioned above was felt to be a drawback  for our purposes. All of these approaches prove upon inspection to center upon analysis of the specific  characteristics of patterns into parts, followed by a  synthesis of the whole from the parts. In these  studies, pattern recognition of the whole, that is, Gestalt recognition, was chosen as a more fruitful  avenue of approach and as a satisfactory problem for  the initial phases of the over-all study." Proceedings of the Eastern Joint Computer Conference, pp. 225-232, New York: Association for Computing Machinery

His present interests are: study of complex equilibria, especially in their topological aspects, as applied to the intelligent and adaptive aspects of the brain. He is now in the Department of Research of Barnwood House Hospital, Gloucester.

The studies reported here have been concerned with the programming of a digital computer to behave in a way which, if done by human beings oranimals, would be described as involving the process of learning. Whilethis is not the place to dwell on the importance of machine-learning procedures,or to discourse on the philosophical aspects,1 there is obviously avery large amount of work, now done by people, which is quite trivial inits demands on the intellect but does, nevertheless, involve some learning.Also in Computers and Thought. Feigenbaum, Edward A. and Julian Feldman (Editors) 1963.See also:IEEE XploreSome Studies in Machine Learning Using the Game of Checkers, II - Recent ProgressIBM Journal of Research and Development, 3:211-229

Frank Rosenblatt, born in New Rochelle, New York, U.S.A., July 11, 1928, graduated from Cornell University in 1950, and received a PhD degree in psychology, from the same university, in 1956. He was engaged in research on schizophrenia, as a Fellow of the U.S. Public Health Service, 1951-1953. He has made contributions to techniques of multivariate analysis, psychopathology, information processing and control systems, and physiological brain models. He is currently a Research Psychologist at the Cornell Aeronautical Laboratory, Inc., in Buffalo, New York, where he Is Project Engineer responsible for Project PARA (Perceiving and Recognizing Automaton). FRANK ROSENBLATT SUMMARY A THEORETICAL brain model, the perceptron, has been developed at the Cornell Aeronautical Laboratory, In Buffalo, New York. The perceptron is a probabilistic system, capable of learning to recognize and differentiate stimuli in its environment. Previous reports have covered the theory of a class of perceptrons ...