Table 1.

A comparison of the previously mentioned studies comparing several characteristics, including their accuracy on the classification task.

Study	Sample (MDDa+HCb)	Electrodes, frequency (Hz)	Preprocessing	Features	MLc models	Accuracy (%)
Ahmadlou et al, 2012 [21]	12+12	7, 256	Wavelets and spectral bands (Fourier), bootstrap	Higuchi and Katz FDd	Enhanced probabilistic neural networks	91.30
Puthankattil and Joseph, 2012 [22]	30 (16 Me+14 Ff)+30	4, 256	Wavelet, total variation filtering, multiresolution decomposition	Wavelet entropy	RWEg, artificial feed forward networks	98.11
Hosseinifard et al, 2014 [24]	45+45	19, 1 kHz	Standard spectral bands	Power, DFAh, Higuchi, correlation dimension, Lyapunov exponent	KNNi, LRj, linear discriminant	90
Faust et al, 2014 [25]	30+30	4 (2 left, 2 right), 256	Wavelet package decomposition	ApEnk, SampEnl, RENm, bispectral phase entropy	PNNn, SVMo, DTp, KNN, NBq, GMMr, Fuzzy Gueno Classifier	99.50
Bairy et al, 2015 [27]	30+30 (left brain only)	N/As	Discrete cosine transform	SampEn, FD, CDt, Hurst exp, LLEu, DFA	DT, KNN, NB, SVM	93.80
Acharya et al, 2015 [26]	15+15	2 left, 2 right, 256	Broadband	FD, LLE, SampEn, DFA, Hv, W-Bxw, W_Byx, EntPhy, Ent1z, DET aa, ENTRab, LAMac, T2 (DDI)ad	SVM, KNN, NB, PNN, DT	98
Mohammadi et al, 2015 [28]	53+43	28 (10/10), 500	Standard bands/FFTae, LDAaf, genetic algorithm	Spectral only	DT	80
Puthankattil and Joseph, 2014 [23]	30+30	4, 256	Wavelet package decomposition	Wavelet entropy, approximate entropy	NNag	98
Liao et al, 2017 [29]	12+12	30, 500	Common spatial pattern	Spectral (common spatial pattern)	KEFB-CSPah	80
Mumtaz et al, 2018 [30]	34/18 F+30/9 Fai	19, 256	RESTaj	Synchronization likelihood	SVM, LR, NB	87.50
Mumtaz et al, 2017 [31]	33+30	19 (EOak, ECal), 256	Fourier	Alpha interhemispheric asymmetry	LR, SVM, NB	98.40
Mumtaz et al, 2018 [32]	34+30	19, 256	10-fold cross-validation	Power, asymmetry, wavelet coefficients, Z-score	LR	94
Bachmann et al, 2018 [35]	13+13	1, 1 kHz	Fourier	HFDam, DFA, Lempel-Ziv complexity, and SASIan	Logistic regression	88
Čukić et al, 2018/2020 [33,34]	26+20	19, 1 kHz	Broadband EEGao, 10-fold cross-validation, PCAap	HFD+SampEn	MPaq, LR, SVM (with linear and polynomial kernel), DT, RFar, NB	97.50

^aMDD: major depressive disorder.

^bHC: healthy control.

^cML: machine learning.

^dFD: fractal dimension.

^eM: male.

^fF: female.

^gRWE: relative wavelet energy.

^hDFA: detrended fluctuation analysis.

ⁱKNN: K-nearest neighbor.

^jLR: linear regression.

^kApEn: approximate entropy.

^lSampEn: sample entropy.

^mREN: Renyi entropy.

ⁿPNN: probabilistic neural network.

^oSVM: support vector machine.

^pDT: decision tree.

^qNB: naïve Bayes.

^rGMM: Gaussian mixture model.

^sN/A: not applicable.

^tCD: correlation dimension.

^uLLE: largest Lyapunov exponent.

^vH: Hurst exponent.

^wW-Bx: higher order spectra features (weighted center of bispectrum [W_Bx]; Acharya et al [26]).

^xW_By: higher order spectra features (weighted center of bispectrum [W_By]; Acharya et al [26]).

^yEntPh: bispectrum phase entropy.

^zEnt1: normalized bispectral entropy.

^aaDET: determinism.

^abENTR: entropy.

^acLAM: laminarity.

^adT2 (DDI): recurrent times.

^aeFFT: fast Fourier transform.

^afLDA: linear discriminant analysis.

^agNN: neural network.

^ahKEFB-CSP: kernel eigen-filter-bank common spatial pattern.

^ai34 depression patients (among them 18 females) and 30 healthy controls (of those 9 were female).

^ajREST: reference electrode standardization technique.

^akEO: eyes opened.

^alEC: eyes closed.

^amHFD: Higuchi fractal dimension.

^anSASI: spectral asymmetry index.

^aoEEG: electroencephalogram.

^apPCA: principal component analysis.

^aqMP: multilayer perceptron.

^arRF: random forest.