Abstract | ||
---|---|---|
The modelling of a dynamic non-linear system is the aim of this work. A neural model is constructed with dynamic artificial neurons, which contain inner feedbacks. Such a model consists of an adder module, a linear dynamic system and a non-linear activation function. Emphasis is put on the searching of the optimal, in the sense of a generalization capability, network architecture. On the basis of a graph of network structures and evaluation values assigned to them, heuristic search algorithms can be elaborated and implemented. The A*-algorithm and Tabu Search are applied and compared. |
Year | DOI | Venue |
---|---|---|
2003 | 10.1007/978-3-540-36562-4_30 | INTELLIGENT INFORMATION PROCESSING AND WEB MINING |
Keywords | Field | DocType |
heuristic search,neural network | Graph,Data mining,Architecture,Heuristic,Adder,Activation function,Computer science,Network architecture,Artificial intelligence,Artificial neural network,Tabu search,Machine learning | Conference |
ISSN | Citations | PageRank |
1615-3871 | 0 | 0.34 |
References | Authors | |
6 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Andrzej Obuchowicz | 1 | 76 | 10.10 |
Krzysztof Patan | 2 | 151 | 18.13 |