[Retracted] Task-Oriented Intelligent Solution to Measure Parkinson’s Disease Tremor Severity

Table 2

Extracted features and their descriptions.

Feature	Domain	Formula
Above mean	T and F
Below mean	T and F
Autocorrelation	T and F
Complexity-invariant distance (CID)	T and F
Sample entropy	T and F
Kurtosis	T and F
Skewness	T and F
Standard deviation	T and F
Max	T and F
Mean	T and F
Median	T and F
Sum of absolute differences (SAD)	T and F
Energy	T and F
Peaks	T
Amplitude of peak PSD	F
Median frequency	F
Frequency dispersion	F
Fundamental frequency	F
Frequency difference	F
Spectral centroid amplitude (SCA)	F
Maximum weighted PSD	F

: window subset contains elements above the mean; : window subset contains elements below the mean; : window length (number of samples); : the acceleration at time ; : the lag. : window’s samples mean; : window’s samples standard deviation; : the probability that two vectors of points within a one window would match; : the probability that two vectors of points within one window would match; : window length is odd; : window length is even; : an element position (index) in the window ; : number of neighbours; : the acceleration at a time ; : the selected window; : the primitive root of unity; : the dispersion frequency in the selected window; : frequency bin; : the lowest frequency in the selected window; : the highest frequency in the selected window; : the range between the median frequency and the lower bound of dispersion frequency, which is equal to the range between median frequency and the higher bound of dispersion frequency, that is, is the range between lower and higher bound of of dispersion frequency; : the at fundamental frequency.