Aiming at the topic of electroencephalogram (EEG) pattern recognition in brain computer interface(BCI), a classification method based on probabilistic neural network (PNN) with supervised learning ispresented in this paper. It applies the recognition rate of training samples to the learning progress of networkparameters. The learning vector quantization is employed to group training samples and the Geneticalgorithm (GA) is used for training the network's smoothing parameters and hidden central vector for determininghidden neurons. Utilizing the standard dataset Ⅰ(a) of BCI Competition 2003 and comparingwith other classification methods, the experiment results show that the best performance of pattern recognitionis got in this way, and the classification accuracy can reach to 93.8 % , which improves over 5 %compared with the best result (88.7 %) of the competition. This technology provides an effective way toEEG classification in practical system of BCI.
Successful restoration of blurred images depends primarily on the knowledge about the degradationparameter.Defocus blur model in the frequency domain is characterized by concentric rings and the blurradius of the point spread function(PSF)can be identified conveniently in the frequency field for peopleby manual means rather than for computer.This paper introduces a practical method for computer to esti-mate the defocus blur parameter in cepstrum area.Fourier transform plays an intermediate role in the pathto cepstrum domain.We suggest a weighted adjustment operation in the frequency domain and then con-vert it to the cepstrum field to increase the accuracy of recognition.
