Time series classification via topological data analysis

In this paper, we develop topological data analysis methods for classification tasks on univariate time series. As an application, perform binary and ternary two public datasets that consist of physiological signals collected under stress non-stress conditions. We accomplish our goal by using persistent homology to engineer stable features after use a delay embedding the subwindowing instead windows fixed length. The combination can be applied any series in application allows us reduce noise long window sizes without incurring extra computational cost. then machine learning models algorithmically engineered obtain higher accuracies with fewer features.