A novel efficient two-phase algorithm for training interpolation radial basis function networks

Huan, Hoang Xuan; Hien, Dang Thi Thu; Huynh, Huu Tuê

doi:10.1016/j.sigpro.2007.05.001

Cited by 8 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Actually, we usually face multivariable interpolation problem and until now, there are still a lot of open problems, which are interested by many people (Bartels et al, 1987). Huan et al proposed a method of producing interpolation function by RBF network, which will be introduced and applied later (Huan et al, 2007).…”

Section: Interpolation Problemmentioning

confidence: 99%

“…This section briefly introduces about RBF network with Gauss radius functions (Bromhead et al, 1988;Haykin, 1999;Huan et al, 2007;Hien, Huan, & Huu-Tue, 2009;Looney, 1997). Technique of Radius Basis Function Powell (Powell, 1988) proposed to calculate interpolation or regression functions by:…”

Section: Rbf Networkmentioning

confidence: 99%

“…Huan et al proposes two-phase algorithm (called HDH algorithm) for RBF interpolation network training with many outstanding advantages in comparison to other popular training algorithms (Huan et al, 2007). When interpolation nodes are equally spaced, HDH algorithm is developed to single-phase algorithm (called QHDH) with much shorter training period and better generality.…”

Section: Rbf Networkmentioning

confidence: 99%

See 2 more Smart Citations

An Effective Solution to Regression Problem by RBF Neuron Network

Hien

Huan

Hoang

2015

International Journal of Operations Research and Information Systems

Self Cite

View full text Add to dashboard Cite

Radial Basis Function (RBF) neuron network is being applied widely in multivariate function regression. However, selection of neuron number for hidden layer and definition of suitable centre in order to produce a good regression network are still open problems which have been researched by many people. This article proposes to apply grid equally space nodes as the centre of hidden layer. Then, the authors use k-nearest neighbour method to define the value of regression function at the center and an interpolation RBF network training algorithm with equally spaced nodes to train the network. The experiments show the outstanding efficiency of regression function when the training data has Gauss white noise.

show abstract

Section: Interpolation Problemmentioning

confidence: 99%

Section: Rbf Networkmentioning

confidence: 99%

Section: Rbf Networkmentioning

confidence: 99%

See 1 more Smart Citation

An Effective Solution to Regression Problem by RBF Neuron Network

Hien

Huan

Hoang

2015

International Journal of Operations Research and Information Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…In their concrete descriptions, MLP realizes a mapping by linearly combining a set of the sigmoidal functions that live on the hidden layer and have global activation domain, and while RBFN does likewise by linearly combining a set of the exponential functions with local activation domain (generally, Gaussian function), as illustrated, respectively, in Fig. 1a, b [13][14][15][16][17]. X In order to implement such networks given a set of limited training data, we can achieve this goal through a learning algorithm in terms of one of supervised, semisupervised, and unsupervised-learning fashions.…”

Section: Introductionmentioning

confidence: 99%

Plane-Gaussian artificial neural network

Yang

Chen

2011

Neural Comput & Applic

View full text Add to dashboard Cite

Multilayer perceptrons (MLPs) and radial basis functions networks (RBFNs) have been widely concerned in recent years. In this paper, based on k-plane clustering (kPC) algorithm, we propose a novel artificial network model termed as Plane-Gaussian network to enlarge the arsenal of the neural networks. This network adopts a so-called Plane-Gaussian activation function (PGF) in hidden neurons. Replacing traditional central point of Gaussian radial basis function (RBF) with central hyperplane, PGF forms a band-shaped rather than spheral-shaped receptive field in RBF, which makes PGF able to express its peculiar geometrical characteristics: locality and globality. Importantly, it is also proved that PGF network (PGFN) having one hidden layer is capable of universal approximation. As a universal approximator, PGFN gives an informal way of bridging the gap between MLP and RBFN. The experiments report comparison between training time and classification accuracies on some artificial and UCI datasets and conclude that (1) PGFN runs significantly faster than MLP and (2) PGFN has comparable or better classification performance than MLP and RBFN, especially in subspace-distributed datasets.

show abstract