Personality Prediction with Hybrid Genetic Programming using Portable EEG Device

Area of Science:

• Neuroscience
• Psychology
• Machine Learning

Background:

• Personality traits influence emotional responses to stimuli.
• Electroencephalogram (EEG) captures brain activity related to emotional states.
• The Myers-Briggs Type Indicator (MBTI) is a widely used personality assessment tool.

Purpose of the Study:

• To develop a real-time method for identifying personality traits using EEG signals.
• To correlate specific brainwave patterns with MBTI personality types.
• To explore the application of machine learning in personality assessment.

Main Methods:

• Utilized a single-channel NeuroSky MindWave 2 for EEG data acquisition.
• Employed Fast Fourier Transform (FFT) for feature extraction from EEG signals.
• Applied Hybrid Genetic Programming (HGP) for classifying EEG data against MBTI traits.

Main Results:

• Developed four two-class HGP classifiers for different MBTI trait groups.
• Achieved an overall classification accuracy of 82.25% using 10-fold cross-validation.
• Demonstrated the feasibility of real-time personality trait identification from EEG.

Conclusions:

• EEG data, analyzed with FFT and HGP, can effectively predict personality traits.
• This approach offers a novel, objective method for personality assessment.
• Future research can refine the model for broader applications in psychology and human-computer interaction.