Title :
Complex dynamic behavior in the evolution of organization with reciprocal functions
Author_Institution :
Inst. of Electron., Acad. Sinica, Beijing, China
Abstract :
The paper presents an experiment in order to exhibit the complex dynamic behavior in the evolution of a function organization with two reciprocal functions. The two functions are realized by multiple layer perceptron, they are modified under external stimuli. Here, evolution is that about the function itself, not about states connected by the function. Under a certain condition, the evolution of two functions exhibit complex behavior like chaos, the feature of dynamic behavior in evolution is changeable following time (here "time" is according to external stimulus).
Keywords :
chaos; dynamics; functional analysis; multilayer perceptrons; chaos; complex dynamic behavior; dynamic behavior; evolution; external stimuli; function organization; multiple layer perceptron; reciprocal functions; Chaos; Equations;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Conference_Location :
Nagoya, Japan
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.714203