An algorithm to generate radial basis function (RBF)-like nets for classification problems

Roy, Asim; Govil, Sandeep; Miranda, R.

doi:10.1016/0893-6080(94)00064-s

Cited by 110 publications

(32 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The results are summarized in table 5. The results as we obtained them are comparable with values reported in literature for BP learning on this dataset: for instance, in a benchmark study by Roy, Govil and Miranda [38] on an earlier version of this dataset containing 608 cases an error-rate of 2.96% was obtained on the test-set, using a multi-layer-perceptron with standard error-backpropagation learning.…”

Section: Other Benchmark Problemssupporting

confidence: 77%

Error-backpropagation in temporally encoded networks of spiking neurons

2002

View full text Add to dashboard Cite

For a network of spiking neurons that encodes information in the timing of individual spike-times, we derive a supervised learning rule, SpikeProp, akin to traditional error-backpropagation and show how to overcome the discontinuities introduced by thresholding. With this algorithm, we demonstrate how networks of spiking neurons with biologically reasonable action potentials can perform complex non-linear classification in fast temporal coding just as well as rate-coded networks. We perform experiments for the classical XOR-problem, when posed in a temporal setting, as well as for a number of other benchmark datasets. Comparing the (implicit) number of spiking neurons required for the encoding of the interpolated XOR problem, it is demonstrated that temporal coding requires significantly less neurons than instantaneous rate-coding.2000 Mathematics Subject Classification: 82C32, 68T05, 68T10, 68T30, 92B20. 1998 ACM Computing Classification System: C.1.3, F.1.1, I.2.6, I.5.1. Keywords and Phrases: spiking neurons; temporal coding; error-backpropagation Note: Work carried out under theme SEN4 "Evolutionary Systems and Applied Algorithmics". This paper has been submitted for publication, a short version has been presented at the European Symposium on Artificial Neural Networks 2000 (ESANN'2000) in Bruge, Belgium. IntroductionDue to its success in artificial neural networks, the sigmoidal neuron is reputed to be a successful model of biological neuronal behavior. By modeling the rate at which a single biological neuron discharges action potentials (spikes) as a monotonically increasing function of inputmatch, many useful applications of artificial neural networks have been build [22; 7; 37; 34] and substantial theoretical insights in the behavior of connectionist structures have been obtained [40; 27].However, the spiking nature of biological neurons has recently led to explorations of the computational power associated with temporal information coding in single spikes [31; 21; 13; 26; 20; 17; 49]. In [32] it was proven that networks of spiking neurons can simulate arbitrary feedforward sigmoidal neural nets and can thus approximate any continuous function. In fact, it has been shown theoretically that spiking neural networks that convey information by individual spike times are computationally more powerful than neurons with sigmoidal activation functions [29].As spikes can be described by 'event' coordinates (place,time) and the number of active (spiking) neurons is typically sparse, artificial spiking neural networks have been shown to allow for very efficient implementations of large neural networks [48; 33]. Single-spike-time computing has also been suggested as a new paradigm for VLSI neural network implementations [28] and would offer a drastic speed-up.Network architectures based on spiking neurons that encode information in the individual spike times have yielded, amongst others, a self-organizing map akin to Kohonen's SOM [39] and a network

show abstract

Section: Other Benchmark Problemssupporting

confidence: 77%

Error-backpropagation in temporally encoded networks of spiking neurons

2002

View full text Add to dashboard Cite

show abstract

“…Where η is the learning rate, M the number of training patterns and ε 0 p,k is the output error, the difference between target and output as in equation (4).…”

Section: Training By Gradient Descentmentioning

confidence: 99%

“…First we find the centers and widths by using some unsupervised clustering algorithm and after that we train the weights among hidden and output units by a supervised algorithm. This process is usually fast [1][2][3][4].…”

Section: Introductionmentioning

confidence: 99%

An Experimental Study on Training Radial Basis Functions by Gradient Descent

Torres-Sospedra

Hernández-Espinosa

Fernández-Redondo

2006

Artificial Neural Networks in Pattern Recognition

View full text Add to dashboard Cite

Abstract. In this paper, we present experiments comparing different training algorithms for Radial Basis Functions (RBF) neural networks. In particular we compare the classical training which consist of an unsupervised training of centers followed by a supervised training of the weights at the output, with the full supervised training by gradient descent proposed recently in same papers. We conclude that a fully supervised training performs generally better. We also compare Batch training with Online training and we conclude that Online training suppose a reduction in the number of iterations.

show abstract

“…For this purpose, there have been proposed various methods to choose RBF centers and widths in batch learning settings [10,11]. However, as far as we know, there is no online version of such an automated learning algorithm.…”

Section: Introductionmentioning

confidence: 99%

An Autonomous Incremental Learning Algorithm for Radial Basis Function Networks

Ozawa¹,

Tabuchi²,

Nakasaka³

et al. 2010

JILSA

Self Cite

View full text Add to dashboard Cite

In this paper, an incremental learning model called Resource Allocating Network with Long-Term Memory (RAN-LTM)is extended such that the learning is conducted with some autonomy for the following functions: 1) data collection for initial learning, 2) data normalization, 3) addition of radial basis functions (RBFs), and 4) determination of RBF centers and widths. The proposed learning algorithm called Autonomous Learning algorithm for Resource Allocating Network (AL-RAN) is divided into the two learning phases: initial learning phase and incremental learning phase. And the former is further divided into the autonomous data collection and the initial network learning. In the initial learning phase, training data are first collected until the class separability is converged or has a significant difference between normalized and unnormalized data. Then, an initial structure of AL-RAN is autonomously determined by selecting a moderate number of RBF centers from the collected data and by defining as large RBF widths as possible within a proper range. After the initial learning, the incremental learning of AL-RAN is conducted in a sequential

show abstract

An algorithm to generate radial basis function (RBF)-like nets for classification problems

Cited by 110 publications

References 17 publications

Error-backpropagation in temporally encoded networks of spiking neurons

Error-backpropagation in temporally encoded networks of spiking neurons

An Experimental Study on Training Radial Basis Functions by Gradient Descent

An Autonomous Incremental Learning Algorithm for Radial Basis Function Networks

Contact Info

Product

Resources

About