Ecg Time Series Classification Via Genetic-Fuzzy Approach Based On Accuracy-Interpretability Trade-Off Optimization

This paper presents the application of our multi-objective-evolutionary-optimization-based (MOEOA-based) design technique of fuzzy rule-based classifiers with genetically optimized accuracy-interpretability trade-off to the problems of ECG time series data classification. First, the ECG200 time series data set coming from the UCR Time Series Classification Archive and used in our experiments is briefly characterized. Then, main components of our approach are outlined. For the purpose of comparison, two MOEOAs are employed in our experiments, i.e., the well-known Strength Pareto Evolutionary Algorithm 2 (SPEA2) and our SPEA2's generalization (referred to as SPEA3) characterized by better performance indices. Our results for the considered ECG time series data are compared with the results of 16 alternative methods, in order to present the advantages (in terms of the optimization of the classifiers' accuracy-interpretability trade-off) of our approach.