Taxonomy of Neural Transfer Functions
نویسندگان
چکیده
The choice of transfer functions may strongly influence complexity and performance of neural networks used in classification and approximation tasks. A taxonomy of activation and output functions is proposed, allowing to generate many transfer functions. Several less-known types of transfer functions and new combinations of activation/output functions are described. Functions parameterize to change from localized to delocalized type, functions with activation based on non-Euclidean distance measures, bicentral functions formed from pairs of sigmoids are discussed.
منابع مشابه
Predicting the Coefficients of Antoine Equation Using the Artificial Neural Network (TECHNICAL NOTE)
Neural network is one of the new soft computing methods commonly used for prediction of the thermodynamic properties of pure fluids and mixtures. In this study, we have used this soft computing method to predict the coefficients of the Antoine vapor pressure equation. Three transfer functions of tan-sigmoid (tansig), log-sigmoid (logsig), and linear were used to evaluate the performance of diff...
متن کاملTransfer functions: hidden possibilities for better neural networks
Sigmoidal or radial transfer functions do not guarantee the best generalization nor fast learning of neural networks. Families of parameterized transfer functions provide flexible decision borders. Networks based on such transfer functions should be small and accurate. Several possibilities of using transfer functions of different types in neural models are discussed, including enhancement of i...
متن کاملSurvey of Neural Transfer Functions
The choice of transfer functions may strongly influence complexity and performance of neural networks. Although sigmoidal transfer functions are the most common there is no a priori reason why models based on such functions should always provide optimal decision borders. A large number of alternative transfer functions has been described in the literature. A taxonomy of activation and output fu...
متن کاملComparison of Three Soft Computing Methods in Estimating Apparent Shear Stress in Compound Channels
Apparent shear stress acting on a vertical interface between the main channel and floodplain in a compound channel serves to quantify the momentum transfer between sub sections of this cross section. In this study, three soft computing methods are used to simulate apparent shear stress in prismatic compound channels. The Genetic Algorithm Artificial neural network (GAA), Genetic Programming (GP...
متن کاملSensitivity analysis of the effective nanofluid parameters flowing in flat tubes using the EFAST method
In the present study, the effective parameters of water-Al2O3 nanofluid flowing in flat tubes are investigated using the EFAST Sensitivity Analysis (SA) method. The SA is performed using GMDH type artificial neural networks (ANN) which are based on validated numerical data of two phase modeling of nanofluid flow in flat tubes. There are five design variables namely: tube flattening (H), flow ra...
متن کامل