New method accurately predicts future behavior of signals with 94.5% accuracy.
The article introduces a new method called TTA to estimate Hurst exponent in signals, which shows how signals behave over time. TTA uses triangles made from three samples to estimate memory in signals. The study tested TTA on synthetic waveforms and found it outperformed existing methods in terms of accuracy, estimation error, computational time, and noise sensitivity. When applied to epilepsy detection, TTA achieved a classification accuracy of 94.5% in distinguishing between different types of EEG signals.
