| During the last ten years, there has been growing interest in the development of Brain Computer Interfaces (BCI). The field has mainly been driven by the needs of completely paralyzed patients to communicate. Confined by situation, most human BCI are based on extracranial electroencephalography (EEG).The BCI system has input (electrophysiological activity from the user), output, and the component which translate input to output. The part of translation consists of feature extraction and translation algorithm, that is, the classifier. This paper is mainly about the classification of EEG by Gaussian classifier and Support Vector Machines (SVM).The training and testing of Gaussian classifier is as follows:At first, "Self-Organizing Map" (SOM) is used to initialize the parameters of the Gaussian classifier. Then, during learning, the initial estimation is improved by gradient descent, and at last we obtain the final parameters. Finally, we test the testing-set. Fixed learning rate is used in this method.Support Vector Machines (SVM) is a multi-dimensional statistical analysis method developed in the 90' s of the 20" century and used to analysis the mutually independent non-gaussian signals. SVM are a kind of novel machine learning methods. It can solve small-sample learning problems better by using Experiential Risk Minimization in place of Structural Risk Minimization. Moreover, this theory can change the problem in non-linearity space to that in the linearity space in order to reduce the algorithm complexity by using the kernel function idea. SVM have become the hotspot of machine learning because of their excellent learning performance. They also have successful applications in many fields, such as: face detection, handwriting digit recognition, text auto-categorization, etc.In this paper, we will make an in-depth study on the basic theory, algorithm and application of SVM with the intention of the EEG pattern recognition. The result and the analysis of it are provided in this paper. |
PDF Full Text Request