Parallel distributed processing: explorations in the microstructure, vol. 2: psychological and biological models
A solution method for truncated normalized max product fuzzy set of equations
International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems
Hi-index | 0.00 |
This paper is concerned with aproposal for a fuzzy artificial neuron with bi-naryinput. The fuzzy neuron is based on fuzzy logic inthat each component of the input vector is compared toa number which represent the membership value for a 0in that position. The results of the comparisons arethen combined using a generalized mean function toproduce a single number which is compared to athreshold as in the case of a perceptron consisting ofa linear combiner with hard limiting function. Atraining algorithm is developed based on an algorithmfor linear inequalities described by Ho and Kashyap ina paper titled ’An Algorithm for Linear Inequalitiesand its Applications‘. The results obtained bysimulation look promising.