Table 2. Feature descriptions.

Feature group	Feature	Description
Amplitude	Mav	mav = mav(signal)
	Mavfd	mavfd = mavfd(signal)
	Mavsd	mavsd = mavsd(signal)
	Peak	peak = max(signal); index(max(signal))
	Rms	rms = rms(signal)
Frequency	Zc	Calculated by comparing each point of the signal with the next; if there is a crossing by zero then it is accounted.
	Fmed	To obtain the median frequency, find the value of the frequency that bisects the area below the X waveform.
	Fmode	This fast Fourier transformation equation is valid for this and the following frequency features:. $X\left(k\right)={\displaystyle {\sum }_{j=1}^{N}x\left(j\right)}{\omega }_N{}^{\left(j-1\right)\left(k-1\right)}$, where ${\omega }_N=e^{\left(-\frac{2\pi i}{N}\right)}$. To find the mode, find the maximum value of X.
	Fmean	$\frac{{\displaystyle {\sum }_{k=1}^{NFFT}X\left(k\right).f\left(k\right)}}{{\displaystyle \sum X\left(k\right)}}$
	Cf	The central frequency is simply the mean of the frequencies that delimit the bandwidth: $cf=\frac{fh-fl}{2}$.
Predictability	Fuzzy entropy	$Saen\left(m,s,d\right)=\ln \left[\frac{C{o}_m\left(s\right)}{C{o}_{m+1}\left(s\right)}\right]$, where m is the window size, s is the similarity standard and d is the signal. It is calculated in a very similar way to the Sample Entropy. The only similarity between the groups is computed by means of a Fuzzy membership function.
	Approximate entropy	For a temporal series with N samples {u(i): 1 ≤ i ≤ N} given m, create vectors $X_j^m$ for each $X_{N-\text{m}+1}^m$ as $X_j^m=\left\{u\left(i\right),u\left(i+1\right),u\left(i+m-1\right)\right\},i=1,\mathrm{\dots },N-m+1$, where m is the number of points to group together for the comparison. For each k ≤ N − m + 1 groups, do $C_k^m\left(r\right)$ which is the number of times the groups had distance less than tolerance r. Then compute the value φm as ${\phi }^m\left(r\right)=\frac{{\displaystyle \sum _{i=1}^{N-m+1}\ln C_j^m\left(r\right)}}{N-m+1}$. The Approximated Entropy is: ApEn(m,r) = limN→∞[φm(r) − φm+1(r)].
Variability	Var	${\sigma }^2=\frac{{\displaystyle \sum _{i=1}^N{\left(x^i-\stackrel{-}{x}\right)}^2}}{N-1}$
	Std	$S=\sqrt{{\sigma }^2}$
	Range	R = MAX(U) − MIN(U)
	Intrange	$SI=\frac{Q3-Q1}{2}$