This paper deals with adaptive control of nonlinear dynamic systems preceded by unknown backlash-like hysteresis nonlinearities, where the hysteresis is described by a dynamic equation. By utilizing this dynamic model and by combining a fuzzy universal function approximator with adaptive control techniques, a stable adaptive fuzzy control algorithm is developed without constructing a hysteresis...

The authors present training and feedforward computation for a single layer of a VMM+WTA classifier. The experimental demonstration of the one-layer universal approximator encourages the use of one-layer networks for embedded low-power classification. The results enabling correct classification of each novel acoustic signal (generator, idle car, and idle truck). The classification structure req...

This paper introduces a no vel quaternion{valued MLP{ type net w ork called the Quaternionic Spinor MLP (QSMLP). In con trast to another recently proposed Quaternionic MLP it uses spinors in its propagation function. This allows very eÆciently the processing of 3D vector data, whic his demonstrated by experiments. The QSMLP is pro ven to be a universal approximator and a learning algorithm for ...

This letter investigates the possibility of using a new universal approximator—support vector machines (SVMs)—as the nonlinear transfer function between oceanic chlorophyll concentration and marine reflectance. The SeaBAM dataset is used to evaluate the proposed approach. Experimental results show that the SVM performs as well as the optimal multilayer perceptron (MLP) and can be a promising al...

We establish upper bounds for the minimal number of hidden units for which a binary stochastic feedforward network with sigmoid activation probabilities and a single hidden layer is a universal approximator of Markov kernels. We show that each possible probabilistic assignment of the states of n output units, given the states of k ≥ 1 input units, can be approximated arbitrarily well by a netwo...

Neural Networks (NN), Type-1 Fuzzy Logic Systems (T1FLS) and Interval Type-2 Fuzzy Logic Systems (IT2FLS) are universal approximators, they can approximate any non-linear function. Recent research shows that embedding T1FLS on an NN or embedding IT2FLS on an NN can be very effective for a wide number of non-linear complex systems, especially when handling imperfect information. In this paper we...

Huang et al. [Universal approximation using incremental constructive feedforward networks with random hidden nodes, IEEE Trans. Neural Networks 17(4) (2006) 879–892] has recently proposed an incremental extreme learning machine (I-ELM), which randomly adds hidden nodes incrementally and analytically determines the output weights. Although hidden nodes are generated randomly, the network constru...