Classification of Interbeat Interval Time-Series Using Attention Entropy

Classification of interbeat interval time-series which fluctuates in an irregular and complex manner is very challenging. Typically, entropy methods are employed to quantify the complexity for classifying. Traditional focus on frequency distribution all observations a time-series. This requires relatively long with at least couple thousands data points, limits their usages practical applications. The also sensitive parameter settings. In this paper, we propose conceptually new approach called attention entropy , pays attention only key observations. Instead counting observations, it analyzes intervals between Attention does not need any tune, robust length, linear time compute. Experiments show that outperforms fourteen state-of-the-art evaluated by real-world datasets. It achieves average classification accuracy AUC = 0.71 while second-best method, multiscale entropy, 0.62 when classifying four groups people length 100.