Qualitative Analysis of Dynamic Activity Patterns in Neural Networks
Ivanka Stamova, Haydar Akça, Gani Tr. Stamov · Journal of Applied Mathematics · 2011
Neural networks have recently been widely used to model some of the human activities in many areas of science and engineering. Mathematical modeling in neural networks has been based on "neurons" that is different both from real biological neurons and from the realistic functioning of simple electronic circuits. These models have received increasing interest due to their impressive applications in areas such as classification, parallel computing, associative memory, pattern recognition, computer vision, and solving some optimization problem. Such applications heavily depend on the dynamic behavior of networks; therefore, the qualitative analysis of these dynamic behaviors is a necessary step for practical design of neural networks.