TY - GEN
T1 - Towards Nature-Inspired Modularization of Artificial Neural Networks via Static and Dynamic Weights
AU - Schuster, Alfons
AU - Berrar, Daniel
PY - 2014/1/1
Y1 - 2014/1/1
N2 - Many conventional artificial neural network (ANN) models are designed for one application domain only. The work presented in this paper describes ANN models that operate with a higher economy by sharing neurons across domains. The use of two different types of weights-static weights and dynamic weights-is a fundamental feature of the presented models. Results from a comprehensive series of experiments provide evidence for the meaningfulness of the proposed models.
AB - Many conventional artificial neural network (ANN) models are designed for one application domain only. The work presented in this paper describes ANN models that operate with a higher economy by sharing neurons across domains. The use of two different types of weights-static weights and dynamic weights-is a fundamental feature of the presented models. Results from a comprehensive series of experiments provide evidence for the meaningfulness of the proposed models.
UR - http://www.scopus.com/inward/record.url?scp=84904759912&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84904759912&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-54121-6_19
DO - 10.1007/978-3-642-54121-6_19
M3 - Conference contribution
AN - SCOPUS:84904759912
SN - 9783642541209
T3 - Communications in Computer and Information Science
SP - 219
EP - 234
BT - Biomedical Informatics and Technology - 1st International Conference, ACBIT 2013, Revised Selected Papers
PB - Springer Verlag
T2 - 1st International Aizu Conference on Biomedical Informatics and Technology, ACBIT 2013
Y2 - 16 September 2013 through 17 September 2013
ER -