Skip to main content

Table 4 A systematic table summarising the machine-learning studies included in the present review, including the features, labels, and algorithms used, as well as the accuracy rates achieved

From: A systematic review of the prediction of consumer preference using EEG measures and machine-learning in neuromarketing research

Citation

Features

Labels

ML algorithms

Accuracy rates

DOI

Aldayel et al. [2]

DEAP dataset, power spectral density and asymmetry

Preference ratings

DNN, KNN, SVM, RF

0.94, 0.88, 0.62, 0.92

https://doi.org/10.3390/app10041525

Aldayel et al. [3]

DEAP dataset, power spectral density

Preference ratings

KNN, SVM, RF, DNN

0.73, 0.81, 0.87, 0.83

https://doi.org/10.3389/fnhum.2020.604639

Al-Nafjan [6]

PCA, mRMR, RFE, ReliefF feature selection algorithms

Like/Dislike self-report responses to Commercial products

DNN, SVM, KNN, LDA, RF

0.93, 0.81, 0.78, 0.70, 0.82

https://doi.org/10.7717/peerj-cs.944

Alimardani and Kaba [4]

Raw EEG signal

Product Choices and movie ratings

CNN, Ensemble Model

0.51–0.75, 0.51–0.64

https://doi.org/10.1145/3460881.3460930

Bandara et al. [23]

PCA feature selection from power spectral density

Movie trailer preference

RF, SVC

91.97, 91.70

https://doi.org/10.1109/ICIIS53135.2021.9660742

Barnett and Cerf [24]

EEG cross-brain correlation

Product willingness to pay

Linear Regression

0.66

https://doi.org/10.1093/jcr/ucw083

Bhushan et al. [31]

TF synchronisation

Binary preference

ANN

74.3

https://doi.org/10.1371/journal.pone.0043351

Gauba et al. [85]

EEG TF effects

Video advertisement preference

RF, DT, Linear regression

0.68, 0.33, 0.041

https://doi.org/10.1016/j.neunet.2017.01.013

Adrián et al. [1]

HR, GSR and EEG TF effects

Video advertisement Valence score

Logistic Regression, SVM, RF

0.66, 0.67, 0.89

https://doi.org/10.3389/fncom.2016.00074

Golnar-Nik et al. [87]

EEG power

Like/dislike ratings

SVM, LDA

0.87, 0.90

https://doi.org/10.1016/j.physbeh.2019.04.025

Guo and Elgendi [101]

TF effects, prepurchase ratings

E-commerce purchase behaviour

Recommender system

N/A

https://doi.org/10.12720/joams.1.1.61–65

Guo et al. [98]

ET, and TF power

Aesthetic preference

SVM, KNN, RF, XGBoost

0.54, 0.61, 0.59. 0.56

https://doi.org/10.1016/j.ergon.2019.02.006

Guixeres et al. [96]

EEG, HRV, ET

Youtube advertisement like/dislike and recall scores

ANN

0.83

https://doi.org/10.3389/fpsyg.2017.01808

Hakim et al. [102]

Frontal band power, hemispheric asymmetry, inter-subject correlations

Product preference

SVM, Logistic regression, DT, KNN,

0.69, 0.67, 0.63, 0.60

https://doi.org/10.1101/317073

Hakim et al. [103]

Frontal band power, hemispheric asymmetry, inter-subject correlations

Product Preference

SVM, Logisitic Regression, KNN, TREE

0.69, 0.67, 0.63, 0.60

https://doi.org/10.1016/j.ijresmar.2020.10.005

Khushaba et al. [139]

Eye-tracking fixation and time–frequency effects

Product preference

Logistic Regression

N/A

https://doi.org/10.1016/j.eswa.2012.12.095

Kumar et al. [151]

Raw EEG signal

Product valence scores

RF

0.74

https://doi.org/10.1016/j.inffus.2018.11.001

Pandey et al. [206]

Discrete Wavelet Transform power and entropy for 5 frequency bands

Movie trailer Likert ratings (rating, familiarity, purchase intent, willingness to spend)

KNN, RF, MP

0.72, 0.71, 0.67

https://arxiv.org/abs/2007.10756

Phutela et al. [210]

Wavelet coefficient, power spectral density, and Hjorth parameters

Advertisement and product preference

NB, SVM, KNN, DT, DL

0.66, 0.66, 0.55, 0.54, 0.55

https://arxiv.org/abs/2206.07484

Ma et al. [178]

ERP

Brand extension acceptance rates

T-SNE algorithm

0.87

https://doi.org/10.3389/fnhum.2021.610890

Mashrur et al. [185]

Frequency, time, and time–frequency metrics over frontal electrodes

Purchase intention and affective attitude towards advertisements

SVM

0.84

https://doi.org/10.3389/fnhum.2022.861270

Mashrur et al. [186]

Frequency, time, and time–frequency metrics over frontal electrodes

Affective attitude towards E-commerce products

SVM

0.94

https://doi.org/10.1016/j.physbeh.2022.113847

Roberts et al. [223]

ERP and eye-movements

Product willingness to pay

Independent component analysis

N/A

https://doi.org/10.3389/fnins.2018.00910

Shestyuk et al. [237]

Frontal asymmetry (alpha/beta), fronto-central power (alpha/theta + theta/gamma)

TV viewership and twitter activity

Linear regression, multiple regression

0.63, 0.72

https://doi.org/10.1371/journal.pone.0214507

Soria Morillo et al. [254]

Frequency band power

Advertisement like/dislike ratings

ANN

0.75

https://doi.org/10.1186/s12938-016–0181-2

Soria Morillo et al. [255]

Frequency band power

Advertisement like/dislike ratings

ANN

0.82

https://doi.org/10.1007/978–3-319–16,480-9_68

Slanzi et al. [251]

Gaze position, pupil dilation, and EEG TF

Website click choice

SVC, logistic regression, neural network

0.69, 0.71, 0.62

https://doi.org/10.1016/j.inffus.2016.09.003

Tyson-Carr et al. [262]

ERP and eye-movements

Product willingness to pay

Independent component analysis

N/A

https://doi.org/10.1016/j.neuroimage.2019.116213

Ullah et al. [263]

Wavelet transformed EEG data

E-commerce product like/dislike ratings

SVM, DT, KNN, ANN

0.80, 0.68, 0.76, 0.81

https://doi.org/10.14569/IJACSA.2022.0130137

Wang et al. [287]

EEG and ET data

Self-reported preference

PCA, Random Lasso, SVM

0.80, 0.75,0.92

https://doi.org/10.1016/j.aei.2020.101095

Wei et al., [291]

EEF TF data

Advertisement effectiveness

SVM

0.75

https://doi.org/10.3389/fnins.2018.00076

Yadava et al. [294]

EEG TF effects

Like/dislike ratings of e-commerce products

HMM

0.68

https://doi.org/10.1007/s11042-017–4580-6

Yılmaz et al. [303]

Power spectral density

Product like/dislike ratings

Logistic regression

N/A

https://doi.org/10.1016/j.cmpb.2013.11.010

Zamani and Naieni [307]

Band power over five brain lobes

Like/dislike ratings of E-commerce products

SVM, ANN, RF

0.87, 0.70, 0.81

https://doi.org/10.18502/fbt.v7i3.4621

Zeng et al. [308]

Power spectral density, hemispheric asymmetry, differential entropy, Hjorth parameters

Sport shoes like/dislike ratings

KNN, SVM

0.94, 0.8

https://doi.org/10.3389/fnhum.2021.793952

Zheng et al. [313]

PSD

Emotion classification

DNN

0.85

https://doi.org/10.1109/TCYB.2018.2797176

Zhu et al. [314]

EEG and ET mixed measures

Self-reported preference

SVM, RF, CNN

70.26, 72.15, 96.4

https://doi.org/10.1016/j.aei.2022.101601