Object
Title: New neural transfer functions
Contributor:
Korbicz, Józef (1951- ) - red. ; Uciński, Dariusz - red.
Group publication title:
Abstract:
The choice of transfer functions in neural networks is of crucial importance to their performance. Although sigmoidal transfer functions are the most common, there is no a-priori reason why they should be optimal in all cases. In this article, advantages of various neural transfer functions are discussed and several new types of functions are introduced. ; Universal transfer functions, parameterized to change from a localized to a delocalized type, are of greatest interest. Biradial functions are formed from products or linear combinations of two sigmoids. Products of N biradial functions in an N-dimensional input space give densities of arbitrary shapes, offering great flexibility in modelling the probability density of the input vectors. ; Extensions of biradial functions, offering a good tradeoff between the complexity of transfer functions and flexibility of the densities they are able to represent, are proposed. Biradial functions can be used as transfer functions in many types of neural networks, such as RBF, RAN, FSM and IncNet. Using such functions and going into the hard limit (steep slopes) facilitates logical interpretation of the network performance, i.e. extraction of logical rules from the training data.
Publisher:
Zielona Góra: Uniwersytet Zielonogórski
Resource Identifier:
Pages:
Source:
AMCS, volume 7, number 3 (1997) ; click here to follow the link
Language:
Rights:
Biblioteka Uniwersytetu Zielonogórskiego
Object collections:
- Repository > Faculties > Faculty of Computer, Electrical and Control Engineering
- Repository > Types of work > Articles
- Repository > Scientific journals and UZ publishing series > International Journal of Applied Mathematics and Computer Science (AMCS)
- Repository > Scientific journals and UZ publishing series > International Journal of Applied Mathematics and Computer Science (AMCS) > International Journal of Applied Mathematics and Computer Science (AMCS) (1997)
Last modified:
Nov 5, 2024
In our library since:
Nov 23, 2020
Number of object content hits:
93
All available object's versions:
https://zbc.uz.zgora.pl/repozytorium/publication/64517
Show description in RDF format:
Show description in OAI-PMH format:
-
International Journal of Applied Mathematics and Computer Science (AMCS)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 26 (2016)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 25 (2015)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 24 (2014)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 23 (2013)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 22 (2012)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 21 (2011)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 20 (2010)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 19 (2009)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 18 (2008)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 17 (2007)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 16 (2006)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 15 (2005)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 13 (2003)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 12 (2002)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 11 (2001)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 10 (2000)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 9 (1999)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 8 (1998)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 7 (1997)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 7, number 1 (1997)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 7, number 2 (1997)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 7, number 3 (1997)
- Boundary stabilization of coupled wave equations
- Optimality conditions for linear 2-D control systems with constraints
- Unreachability and uncontrollability of 2-D linear systems with bounded inputs
- Sub-gradient algorithms for solving multi-dimensional analysis problems of dissimilarity data
- Gradient of the discretized energy method and discretized continuous gradient in electromagnetic shaping simulation
- Genetic algorithms and Hopfield neural networks for solving combinatorial problems
- Kinematics of free-floating robots revisited
- Fuzzy control of robotic manipulators
- New neural transfer functions
- Universal voice compression algorithms based on approximate string matching
- Computing in GF(2m) using GAP
- Towards similarities of interpretations of temporal logic formulae
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 7, number 4 (1997)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 7, number 1 (1997)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 6 (1996)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 5 (1995)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 4 (1994)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 3 (1993)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 2 (1992)
-
International Journal of Applied Mathematics and Computer Science (AMCS),volume 1 (1991)
-
International Journal of Applied Mathematics and Computer Science (AMCS), volume 26 (2016)
| Edition name | Date |
|---|---|
| New neural transfer functions | Nov 5, 2024 |
