[Retracted] Review on Epileptic Seizure Prediction: Machine Learning and Deep Learning Approaches

Table 2

Most prominent features that are used for EEG seizure detection.

Method used	Features extracted
Analysis in time	Mean, skewness, kurtosis, entropy, median, mode, entropy, fuzzy entropy, Hurst exponent, variance, max, min, zero crossings, line length, energy, power, Shannon entropy, sample entropy, approximate, and standard deviation
Analysis in frequency	Spectral energy, peak frequency, median frequency, spectral power, and spectral entropy
Time/frequency combination	Line length, min, max, standard deviation, energy, median, Shannon entropy, approximate entropy, and root mean square
Wavelet analysis	Variation, bounded variation, relative power, relative scale energy, coefficients, energy, entropy, relatively bounded, variance, and standard deviation