Feasibility of real-time calculation of correlation integral derived statistics applied to EEG time series (Q556378)

A framework is presented on how to speed up the computation of the correlation integral \(c(r)\). It is described how to optimize the benefit for the information overlap between successive distances rather than between overlaping epochs. The suggested framework presents an introduction of a nonrandom instead of a random distance exclusive scheme. The advantage of the nonrandom is the possibility of regaining information from preceding distance calculations. I think that (1) If ranges of parameters \(p, s,\) etc. were extended, also parameter considerations, conditions, then the results may be different; (2) A simulation study is important to support the result in this paper.