“…While in one hand, the choices are profoundly based on the input features that are considered based on the morphological features garnered from the time domain [8], [19], [20], [29], complex heartbeat representations [21], wavelet transforms [9], [13], [14], [15], [16], [23], [30], or higher order statistics (HOS) [9], [11], [12], [14] and frequency-domain features [11], [12], [21], [22], [31]. In the other dimension, the feature selection models like the independent component analysis (ICA) [23], [30], PSO (particle swarm optimization) [21] and the GA-BPNN (genetic algorithm back propagation neural networks) are used.…”