Abstract

A massively parallel neural network architecture, called a masking field, is characterized through systematic computer simulations. A masking field is a multiple-scale self-similar automatically gain-controlled cooperative–competitive feedback network F₂. Network F₂ receives input patterns from an adaptive filter F₁ → F₂ that is activated by a prior processing level F₁. Such a network F₂ behaves like a content-addressable memory. It activates compressed recognition codes that are predictive with respect to the activation patterns flickering across the feature detectors of F₁ and competitively inhibits, or masks, codes which are unpredictive with respect to the F₁ patterns. In particular, a masking field can simultaneously detect multiple groupings within its input patterns and assign activation weights to the codes for these groupings which are predictive with respect to the contextual information embedded within the patterns and the prior learning of the system. A masking field automatically rescales its sensitivity as the overall size of an input pattern changes, yet also remains sensitive to the microstructure within each input pattern. In this way, a masking field can more strongly activate a code for the whole F₁ pattern than for its salient parts, yet amplifies the code for a pattern part when it becomes a pattern whole in a new input context. A masking field can also be primed by inputs from F₁: it can activate codes which represent predictions of how the F₁ pattern may evolve in the subsequent time interval. Network F₂ can also exhibit an adaptive sharpening property: repetition of a familiar F₁ pattern can tune the adaptive filter to elicit a more focal spatial activation of its F₂ recognition code than does an unfamiliar input pattern. The F₂ recognition code also becomes less distributed when an input pattern contains more contextual information on which to base an unambiguous prediction of which the F₁ pattern is being processed. Thus a masking field suggests a solution of the credit assignment problem by embodying a real-time code for the predictive evidence contained within its input patterns. Such capabilities are useful in speech recognition, visual object recognition, and cognitive information processing. An absolutely stable design for a masking field is disclosed through an analysis of the computer simulations. This design suggests how associative mechanisms, cooperative–competitive interactions, and modulatory gating signals can be joined together to regulate the learning of compressed recognition codes. Data about the neural substrates of learning and memory are compared to these mechanisms.

© 1987 Optical Society of America

Probing cognitive processes through the structure of event-related potentials during learning: an experimental and theoretical analysis

Jean-Paul Banquet and Stephen Grossberg
Appl. Opt. 26(23) 4931-4946 (1987)

ART 2: self-organization of stable category recognition codes for analog input patterns

Gail A. Carpenter and Stephen Grossberg
Appl. Opt. 26(23) 4919-4930 (1987)

Neural dynamics of attentionally modulated Pavlovian conditioning: blocking, interstimulus interval, and secondary reinforcement

Stephen Grossberg and Daniel S. Levine
Appl. Opt. 26(23) 5015-5030 (1987)

Adaptive bidirectional associative memories

Bart Kosko
Appl. Opt. 26(23) 4947-4960 (1987)

Adaptive, associative, and self-organizing functions in neural computing

Teuvo Kohonen
Appl. Opt. 26(23) 4910-4918 (1987)