Q. Zhang, Using Wavelet Network in Non-parametric Estimation, IEEE Trans. on Neural Network, Vol. 8, No. 2, March 1997, pages 227-236.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....filtering [7] or an AR (autoregressive) model [14] See also [8] which relates the wavelet transform to a multiscale autoregressive type of transform. Wavelet networks are neural networks (supervised mapping networks) with wavelet functions replacing the more usual sigmoid transfer functions [19, 20]. 2 Wavelets for Feature Discovery Our task is to consider the approximation of a time series at coarser and coarser resolution, summarized in a multiresolution decomposition. The individual time series resulting from the decomposition, taken together, can provide a detailed picture of the ....

Q. Zhang, Using wavelet network in nonparametric estimation, IEEE Transactions on Neural Networks 8, 227--236, 1997.

.... presented in the neural network framework [58] RBFNs were reintroduced as a particular case of regularization networks [68] Support Vector Machines [23] reintroduced RBFNs again in the framework of Statistical Learning Theory and Kernel Based Algorithms [59] The architecture of Wavelet Networks [91] is a particular case of RBFNs. Independently, the fuzzy logic community developed fuzzy controllers [6] whose e#ectiveness rely on the same approximation principles [44] Closely related to the fuzzy approach some research [11,79,80] proposed to use the RBFN for mapping and refining propositional ....

....to other machine learning techniques in the task of diagnosis of pigmented skin lesions. Theoretical bounds on the generalization error can improve confidence in the system and its chances of acceptance. 7 RBFNs as Wavelet Networks Wavelet Networks (WN) were proposed as nonparametric regressors [92,91]. The proposal relies on the results of the broad area of wavelet theory and wavelet analysis that is very popular for signal analysis and data compression. In a nutshell, the basic idea of the wavelet transform is to analyse signals in terms of local variability with more flexibility than the ....

Q. Zhang. Using wavelet network in nonparametric estimation. IEEE Transactions on Neural Networks, 8(2):227--236, March 1997.

....are the attempts to apply techniques studied for a particular kind of network to a different paradigm. On the other hand, it has been recently pointed out [1] 2] that most of the neural paradigms, such as Multi Layer Perceptron (MLP) Radial Basis Function (RBF) networks and Wavelet Networks (WN) [3], can be viewed under a unified perspective by means of the Weighted Radial Basis Function (WRBF) paradigm. One of the most important consequences of unification is that some initialization and training procedures studied for one particular neural paradigm could be applied with slight ....

....: K) is given, which should be fit by the model and, due to the high nonlinearity in the input output relation (1) the search for the optimal parameter values require iterative numerical procedures. The idea which lays behind the OLS algorithm (introduced in [4] for RBF networks and adopted in [3] for WNs) is that, once m is chosen and w j , c j and b j are fixed, the model (1) depends linearly on the parameters a j , which can therefore be determined by the standard Least Square (LS) method. Therefore the problem is essentially divided in three steps: ffl the construction of a library of ....

[Article contains additional citation context not shown here]

Q. Zhang: "Using Wavelet Network in Nonparametric Estimation," IEEE Transactions on Neural Networks, Vol. 8, No. 2, pp. 227--236, March 1997.

....engineering, economics, agriculture, technology, etc. there exist several modeling, forecasting and classification problems where soft computing and non linear statistic approaches provide advantages with respect to other methods. Among other soft computing methods, neuro wavelet networks (NWN s) [1, 2, 3] seem to cope well with such problems, as they are good approximators of strongly non linear functions. Furthermore they do not need any a priori assumption on the kind of relationships linking input and output variables, due to their capability of automatically learning from examples [2] When ....

Q. Zhang: "Using Wavelet Network in Non-parametric Estimation," IEEE Transactions on Neural Networks, Vol. 8, No. 2, pp. 227-236, March 1997.

....of such almost orthogonal basis functions. In the general case where the basis functions in G are not orthogonal, in order to overcome the combinatorial complexity of the exhaustive search, three different heuristics are reviewed in the following, details of these algorithms can be found in (Zhang, 1994). The residual based selection (RBS) The idea of this method is to select, for the first stage, the basis function in G that best fits the estimation data, then repeatedly select the basis function from the remainder of G that best fits the residual of the previous fitting. In the literature of ....

....expensive when G is large. Continuous wavelet transform in combination with basis function selection Applying the above mentioned techniques of basis function selection to non orthogonal wavelets yields an interesting family of models called wavelet networks (Zhang and Benveniste, 1992; Zhang, 1994). Though they are computationally less efficient than the wavelet shrinking algorithms in small dimensional case, they allow to handle problems of moderately large dimensions. The software package of wavelet networks in Matlab language is available via anonymous FTP (Zhang, 1993) We need to ....

[Article contains additional citation context not shown here]

Zhang, Q. (1994). Using wavelet network in nonparametric estimation. Technical Report 833, IRISA.

No context found.

Q. Zhang, Using Wavelet Network in Non-parametric Estimation, IEEE Trans. on Neural Network, Vol. 8, No. 2, March 1997, pages 227-236.

No context found.

Q. Zhang. Using wavelet network in nonparametric estimation. IEEE Trans. Neural Networks, 8(2):227{ 236, 1997.

No context found.

Q. Zhang, "Using Wavelet Network in Nonparametric Estimation", INRIA, Technical Report No. 2321, June 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC