摘要 :
Based on the least--square minimization a computationally efficient learning algo- rithm for the Principal Component Analysis(PCA) is derived. The dual learning rate parameters are adaptively introduced to make the proposed algori...
展开
Based on the least--square minimization a computationally efficient learning algo- rithm for the Principal Component Analysis(PCA) is derived. The dual learning rate parameters are adaptively introduced to make the proposed algorithm providing the capability of the fast convergence and high accuracy for extracting all the principal components. It is shown that all the information needed for PCA can be completely represented by the unnormalized weight vector which is updated based only on the corresponding neuron input--output product. The convergence performance of the proposed algorithm is briefly analyzed. The relation between Oja's rule and the least squares learning rule is also established. Finally, a simulation example is given to illustrate the effectiveness of this algorithm for PCA.
收起
摘要 :
An efficient method is proposed to design 4-band scaling functions with the following five advan- tages: compact support, orthogonality, symmetry, regularity, and interpolation; and a family of such scaling functions with the shor...
展开
An efficient method is proposed to design 4-band scaling functions with the following five advan- tages: compact support, orthogonality, symmetry, regularity, and interpolation; and a family of such scaling functions with the shortest support is given.
收起
摘要 :
This paper starts with the discussion of the principle of Reduced-Rank (RR) Space- Time Adaptive Processing (STAP). It is followed by a dedication of the upper bound performance of all eigen--based RR methods provided by Cross Spe...
展开
This paper starts with the discussion of the principle of Reduced-Rank (RR) Space- Time Adaptive Processing (STAP). It is followed by a dedication of the upper bound performance of all eigen--based RR methods provided by Cross Spectral Method (CSM) under the condition of a given processor rank and an identical secondary sample size. A performance comparison between two RR STAP processors with prefixed structure and CSM is performed by the means of simulations. It is shown that the performance of time pre--filtering followed by jointly localized STAP structure (i.e. 3DT-SAP) is very close to the upper bound and thereby it is an effective RR approach.
收起
摘要 :
By use of the approach of complex random signal processing, the asymptotic statis- tical properties of the least square estimates of 2-D exponential signals are studied. In doing so it is found that the representation is considera...
展开
By use of the approach of complex random signal processing, the asymptotic statis- tical properties of the least square estimates of 2-D exponential signals are studied. In doing so it is found that the representation is considerably more intuitive, and is analytically more tractable.
收起
摘要 :
In recent years, M-band orthonormal wavelet bases, due to their good character- istics, have attracted much attention. The ability of 2-band wavelet packets to decompose high frequency channels can be employed to improve the perfo...
展开
In recent years, M-band orthonormal wavelet bases, due to their good character- istics, have attracted much attention. The ability of 2-band wavelet packets to decompose high frequency channels can be employed to improve the performance of wavelets for time-frequency localization, which makes more kinds of signals for analyzing by wavelets. Similar to the nota- tions from the extension of 2-band wavelets to 2-band wavelet packets, the theoretic framework of M-band wavelet packets is developed, a generalization of the notations and properties of 2-band Wavelet packets to that of M-band wavelet packets is made the corresponding proofs are Given.
收起
摘要 :
In this paper, a new method to solve multiscale difference equation(MSDE) with the M-band wavelet neural networks is proposed. It is show that the method has many advantages Over the existing methods and enlarges the range of the ...
展开
In this paper, a new method to solve multiscale difference equation(MSDE) with the M-band wavelet neural networks is proposed. It is show that the method has many advantages Over the existing methods and enlarges the range of the solvable equations.
收起
摘要 :
A neural network for implementing the Marcos 'propagator' bearing estimation algorithm is presented. It is shown both analytically and by simulations that this neural network is guaranteed to be stable and to provide the results a...
展开
A neural network for implementing the Marcos 'propagator' bearing estimation algorithm is presented. It is shown both analytically and by simulations that this neural network is guaranteed to be stable and to provide the results arbitrarily close to the accurate propagator operator and the orthogonal projection operator on the noise subspace within an elapsed time of only a few characteristic time constants of the network. The parameters such as the interconnection strengths and bias currents of this proposed network can be obtained from the cross-spectral matrix without any computations.
收起
摘要 :
The authors construct a three-band compactly supported orthogonal scaling function with an integer-shifted sampling property. Waiter's sampling theorem for wavelet subspaces corresponding to this scaling function has a compactly s...
展开
The authors construct a three-band compactly supported orthogonal scaling function with an integer-shifted sampling property. Waiter's sampling theorem for wavelet subspaces corresponding to this scaling function has a compactly supported interplant. Therefore, the signals in multiresolution spaces can be reconstructed quickly and accurately without any truncation errors.
收起
摘要 :
It is shown that the number of vanishing moments of the even number-order orthogonal M-band compactly supported interpolating scaling function increases by one. Consequently. The asymptotic performance of the Mallat projection is ...
展开
It is shown that the number of vanishing moments of the even number-order orthogonal M-band compactly supported interpolating scaling function increases by one. Consequently. The asymptotic performance of the Mallat projection is the same as the orthogonal projection for the L/sub 2/-approximation of smooth functions.
收起
