Nonparametric statistics in multivariate time series for cognitive anomaly detection
Vladimir L. Gorokhov, Dmitry V. Kholodnyak · 2016
The algorithms of exposure of signals of unknown form are offered in multidimensional time series with subsequent presentation of separate time series from every mining hole as points multidimensional space of parameters of these mining holes. By means of facilities of cognitive machine graphic arts these mining holes form κ cognitive characters in that evidently as aberrant behavior of parameters of these mining holes can be educed in case of multidimensional terabyte of time series.