Analysis Of Evaluating Node Importance Based On Complex EEG Network

In recent years, the study of complex networks has developed in various fields, as one research field of complex network, the analysis of evaluating node importance has become an essential role in this area. Identification of key nodes has high application value in the reality, the networks’ reliability can be improved through perfecting their key nodes, and on the other hand, the research can destroy some networks by attacking these nodes deliberately.Recently, there are many kinds of researches to find the key nodes in networks, which can be divided into two kinds, one kind is the node important degree is equal to significant, another class of the node important degree is equal to the degree of destruction. To improve the performance, TOPSIS analysis is introduced to combine these essential measures and consider the fact that the network is changing constantly.Given full consideration to the brain network topology characteristics, the α wave、β wave、δ wave、and θ wave of the brain cranial nerve are analyzed by using Hibert phase synchronous analysis in this thesis. First of all, the 64 electrodes of electrode cap are identified as network nodes; secondly, the data are carried out offline; phase matrix is gotten by using the Hilbert phase synchronization analysis method; proper threshold is set to get the brain functional network.The degree and degree distribution 、 clustering coefficient 、 average length 、 closeness、 betweenness and so on are worked out by using MATLAB. And then node importance is evaluated by using TOPSIS. And then by contrast with node topology potential method which can identify the node, the method of TOPSIS is proved to have high recognition. The TOPSIS is proved to be effective on finding out key nodes. The changing of the brain network can be known intuitively by drawing brain electrical network diagram in every state using Pajek.The key nodes are always evaluated by using single measure. The correlation between them is ignored. It don’t accord with real fact. The networks can be effected by many factors, so it should be considerated conprehensively. TOPSIS is adopted to approach the ideal results. The effective parameter index is integreted to identify the key node for the brain, which can analyze the change of the barin networks from the perspective of informatics in this thesis. A new method to find the complex networks’ key nodes is provided by TOPSIS.