Research On The Disease Diagnosis Expert System By Using Nonlinear Dynamic Parameters Of High Frequency EEG In Hadoop Cloud Environment

The high frequency EEG is the frequency at 36-44 Hz paroxysmal EEG.The high frequency EEG carries a wealth of thinking activity information,and is a reflection of the status and function of the brain.Based on the research of the high frequency electrical to reflect the attention,image thinking and abstract the mechanism of mental activity.High frequency EEG signal with a wide range of non-linear dynamics is chaotic.On the basis of the further study of contrast to choose two classification effect good nonlinear dynamic parameters of the correlation dimension and Lyapunov index of high-frequency EEG data for processing,Nonlinear dynamic characteristic parameters of extracting high-frequency EEG data classification judgment.The calculation method of correlation dimension and Lyapunov exponent was the traditional G-P and the small amount of data.Designed and trained the RBF probabilistic neural network by using the principles of RBF neural network.Constructed the RBF probabilistic neural network based on correlation dimension and Lyapunov exponent,the network was used as the expert system reasoning basis.Results show that the RBF neural network classification effect and the learning speed are better than BP neural network algorithm.For real time,continuous,lead more people,the complexities of disease prevention and monitoring.The research on application of cloud computing the distributed parallel computing technology,responsive to the real-time data processing parallelism and also a large amount of data storage.On the basis of this study,this research adopts the Hadoop distributed storage of cloud computing architecture and parallel distributed computing technology was designed and implemented based on Hadoop cloud environment probability of high-frequency EEG nonlinear dynamic parameters of RBF neural network aided diagnosis expert system prototype.