A Preliminary Study on Human Trust in Pseudo-Real-Time Scenario through Electroencephalography and Machine Learning based Data Classification

• DOI: 10.46254/ev01.20230061
• Publication Date: 2024-01-08T20:02:56Z
• Authors: Kazi Farzana Firoz; Younho Seong; Sun Yi; Yoo-Sang Chang

This study aims to sense trust and distrust in a real-time inspired scenario through the classification of brain signals.Here, word elicitation is used invoke mental state associated with machine.Participants think any event or experience that comes into their mind when they observe word.They recall event/experience without deliberately filtering out kind cognitive affective state, which we consider as replica real-life where all kinds states emotions possibly co-exist along distrust.While thinking recalling such events, Electroencephalography data recorded from participants' cortex analyzed Machine Learning approaches several algorithms.The developed an approach whether human going compared different methods discuss efficacies scenarios.Here, individualistic generalistic delved into, it found provide better accuracy sensing brain.Also, this explored ways increase efficiency method by reducing number channels performance models observing loss caused reduced channels.This K-Nearest Neighbor and/or Random Forest classifier algorithm provides best result using raw most scenarios, achieving up 100% average accuracy.