Calculation of EEG correlation dimension: large massifs of experimental data

Correlation dimension of reconstructed attractor (D(2)) is one of the specific values for human electroencephalogram (EEG). It makes it possible to evaluate variability of human brain functioning. There are some requirements, made for time series to be analyzed by Grassberger-Procaccia method, and EEG does not meet them perfectly. Also, realization of this algorithm uses subjective evaluation procedure and needs setting of additional parameters for calculation, such as lag and embedding dimension. Method of calculation, suggested in the article, meets specifics of EEG and allows calculating without using subjectivity. This makes it possible to process large amounts of experimental data. The value, calculated by this method, is not D(2) in strict sense, but it is useful for evaluated variability of brain functioning.