Discriminant analysis of HRV´s autoregressive model coefficient for cardiac arrhythmia

An attempt is made to obtain an autoregressive (AR) model coefficient for heart rate variability (HRV) and then to classify the symptom groups of the coefficient. As a result of analyzing 104 sets of clinical data, the following conclusions are drawn. (1) HRV obtained from clinical data can be considered time-series data. (2) The optimal condition for classification is to assume that the coefficient is at 110 R-R intervals with 11th order. This fixed order shows a tendency to approach the largest one among the representative optimal orders of each symptom