Fuzzy ART properties

Huang, Juxin; Georgiopoulos, Michael; Heileman, Gregory L.

doi:10.1016/0893-6080(94)00073-u

Cited by 75 publications

(30 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Also, due to the specific nature of their neural architecture, responses of FA and FAM to specified inputs are easily explained (Carpenter & Tan, 1995), in contrast to other neural network models, where, in general, it is more difficult to explain why an input pattern x produced an output y. Properties of learning for FA and FAM can be found in their original references (Carpenter et al, 1991(Carpenter et al, , 1992, as well as in the work of others (Georgiopoulos, Fernlund, Bebis, & Heileman, 1996;Huang, Georgiopoulos, & Heileman, 1995).…”

Section: Introductionmentioning

confidence: 99%

Category regions as new geometrical concepts in Fuzzy-ART and Fuzzy-ARTMAP

Anagnostopoulos¹,

Georgiopoulos²

2002

Neural Networks

Self Cite

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Category regions as new geometrical concepts in Fuzzy-ART and Fuzzy-ARTMAP

Anagnostopoulos¹,

Georgiopoulos²

2002

Neural Networks

Self Cite

View full text Add to dashboard Cite

“…In this section we analyze the effect of a variable vigilance on the category learning process and on the self-stabilization of fuzzy ART. Self-stabilization with fixed vigilance has been documented at length [3], [6], [11].…”

Section: A Discrimination and Generalizationmentioning

confidence: 99%

“…This prevents the creation of an infinite number of categories, or perpetual category updates, upon repeated presentations of some inputs. This definition has been used for analyzing fuzzy ART's stability for a fixed vigilance [3], [11], [12], [18].…”

Section: B Stable Category Learning With Variable Vigilancementioning

confidence: 99%

“…Since extensive literature exists on fuzzy ART and other ART models [3], [4], [6], [8]- [11], only the algorithmic aspects pertinent to multiple categorization are reviewed. In Section III, a "tunable" choice function is introduced, and the resulting order of search through the categories is compared to that of the original choice function.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Generalization, discrimination, and multiple categorization using adaptive resonance theory

Lavoie¹,

Crespo²,

Savaria

1999

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

Abstract-The internal competition between categories in the adaptive resonance theory (ART) neural model can be biased by replacing the original choice function by one that contains an attentional tuning parameter under external control. For the same input but different values of the attentional tuning parameter, the network can learn and recall different categories with different degrees of generality, thus permitting the coexistence of both general and specific categorizations of the same set of data. Any number of these categorizations can be learned within one and the same network by virtue of generalization and discrimination properties. A simple model in which the attentional tuning parameter and the vigilance parameter of ART are linked together is described. The self-stabilization property is shown to be preserved for an arbitrary sequence of analog inputs, and for arbitrary orderings of arbitrarily chosen vigilance levels.

show abstract

“…The choice parameter [7] provides a floating point overflow, if . In [9] some additional properties of Fuzzy ART with variations on are pointed out, such as, e.g., lowest possible vector size of prototypes. Simulations in this paper are performed with a value of .…”

Section: Referring Tomentioning

confidence: 99%