Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 1, 20192019 International Seminar on Application for Technology of Information and Communication (iSemantic)

Emotion classification using a Naïve Bayes classifier on combined alpha and beta EEG frequency bands achieved an accuracy of 87.5% for differentiating happy and sad emotions.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Emotion recognition using physiological signals is of interest in human-computer interaction, with EEG considered the most reliable modality for understanding emotion processing and perception.

Population

12 participants

Comparison

Happy vs sad emotion classification using time domain features from alpha and beta EEG frequency bands

Key result

Emotion classification using a Naïve Bayes classifier on combined alpha and beta EEG frequency bands achieved an accuracy of 87.5% for differentiating happy and sad emotions.

Authors

NONur Yusuf OktaviaAWAdhi Dharma WibawaEPEvi Sentiana Pane

Discussion

Loading...

Member takes

Overview

Hypothesis-generating for EEG-based emotion detection; leaves open clinical translation pending larger validation studies.

Structured PICO

P
Population
12 participants providing an EEG based emotion dataset with 4 recording channels (AF3, AF4, O1, O2)
I
Intervention
Naïve Bayes learning classifier using time domain features (mean, standard deviation, number of peaks) extracted from alpha and beta frequency bands of EEG signals
O
Outcome
Accuracy of emotion recognition (differentiating happy and sad)

A Naïve Bayes classifier using combined alpha and beta band time-domain features from EEG signals achieved 87.5% accuracy in classifying happy versus sad emotions.

Cite This Study

Oktavia et al. (2019) studied Emotion recognition (n=12). Naïve Bayes learning classifier on EEG signals was evaluated on Accuracy of emotions recognition. Emotion classification using a Naïve Bayes classifier on combined alpha and beta EEG frequency bands achieved an accuracy of 87.5% for differentiating happy and sad emotions.

synapsesocial.com/papers/6a17b0c640149b897cb441d1https://doi.org/10.1109/isemantic.2019.8884224
View Full Paper
Ask AI
Bookmark
Share