University of Dortmund
نویسندگان
چکیده
A multi-layered neural network based on multi-valued neurons is considered in the paper. It is shown that using a traditional architecture of multi-layered feedforward neural network (MLF) and the high functionality of the multi-valued neuron, it is possible to obtain a new powerful neural network. Its training does not require a derivative of the activation function and its functionality is higher than the functionality of MLF containing the same number of layers and neurons. These advantages of MLMVN are confirmed by testing using Parity n, two spirals and "sonar" benchmarks.
منابع مشابه
Towards Decentralized Production: A Novel Method to Identify Flexibility Potentials in Production Sequences Based on Flexibility Graphs
E-mail: [email protected] ∗2Institute of Machine Tools and Manufacturing, ETH Zurich, Zurich, Switzerland ∗3Chair of Enterprise Logistics, Technical University of Dortmund, Dortmund, Germany ∗4Department of Computer Science VII, Technical University of Dortmund, Dortmund, Germany ∗5F/L/S Fuzzy Logik Systeme GmbH, Dortmund, Germany ∗6Fraunhofer Institute for Material Flow and...
متن کاملModified repeated median filters
T. BERNHOLT, R. FRIED, U. GATHER and I. WEGENER a Department of Computer Science, University of Dortmund, 44221 Dortmund, Germany [email protected] [email protected] b Department of Statistics, University Carlos III de Madrid, 28903 Getafe, Spain [email protected] c Department of Statistics, University of Dortmund, 44221 Dortmund, Germany [email protected]...
متن کاملPredictions for Neutrino Structure Functions
S. Kretzer, a,b F.I. Olness, c R.J. Scalise, c R.S. Thorne, d∗ U.K. Yang Univ. Dortmund, Dept. of Physics, D-44227 Dortmund, Germany Michigan State University, Dept. of Physics and Astronomy, East Lansing, MI 48824-1116 Southern Methodist University, Department of Physics, Dallas, TX 75275-0175 Cavendish Laboratory, University of Cambridge, Madingley Road, Cambridge, CB3 0HE University of Chica...
متن کاملGermanTeam 2003
1 Lehrstuhl für Datenverarbeitungssysteme, FB Elektrotechnik und Informationstechnik, University of Dortmund, Otto-Hahn-Strasse 4, 44221 Dortmund, Germany 2 Institut für Informatik, LFG Künstliche Intelligenz, Humboldt-Universität zu Berlin, Rudower Chaussee 25, 12489 Berlin, Germany 3 Fachgebiet Simulation und Systemoptimierung, FB 20 Informatik, Technische Universität Darmstadt, Alexanderstra...
متن کاملCluster Analytic Strategy for Identification of Metagenes Relevant for Prognosis of Node Negative Breast Cancer
Worldwide, breast cancer is the second leading cause of cancer deaths in women. To gain insight into the processes related to the course of the disease, human genetic data can be used to identify associations between gene expression and prognosis. Moreover, the expression data of groups of genes may be aggregated to metagenes that may be used for investigating complex diseases like breast cance...
متن کاملFurther evidence for asymptotic safety of quantum gravity
K. Falls, D. Litim, K. Nikolakopoulos, and C. Rahmede Department of Physics and Astronomy, University of Sussex, Brighton BN1 9QH, United Kingdom Institute for Theoretical Physics, University Heidelberg, 69120 Heidelberg, Germany Karlsruhe Institute of Technology, Institute for Theoretical Physics, 76128 Karlsruhe, Germany Technical University Dortmund, Institute for Theoretical Physics, 44221 ...
متن کامل