Classification of EEG Signals Using Neural Network for Predicting Consumer Choices

Area of Science:

• Neuroscience and Biomedical Engineering
• Signal Processing and Machine Learning

Background:

• Electroencephalogram (EEG) records brain activity, offering insights into neurological processes and diseases.
• EEG signals are weak, susceptible to noise, necessitating robust preprocessing for accurate analysis.
• Understanding brainwave patterns is vital for diagnosing conditions like Alzheimer's and brain damage.

Purpose of the Study:

• To investigate methodologies for preprocessing and analyzing EEG data.
• To develop predictive models for analyzing consumer preferences using EEG signals.
• To explore the application of EEG in human-computer interaction, specifically in understanding online consumer behavior.

Main Methods:

• Acquisition of EEG signals from individuals performing online searches for various products.
• Implementation of preprocessing techniques to clean and enhance the weak EEG signals.
• Development and application of predictive modeling for classifying consumer likes and dislikes based on EEG data.

Main Results:

• Demonstrated the feasibility of using EEG signals to infer consumer preferences during online shopping.
• Highlighted the importance of signal preprocessing for reliable EEG data interpretation.
• Showcased the potential of EEG in novel applications like human-computer interaction and market research.

Conclusions:

• EEG analysis, with appropriate preprocessing, can effectively capture subtle brain responses related to consumer choices.
• This research opens avenues for advanced human-computer interfaces and personalized online experiences.
• Further studies can refine EEG-based predictive models for broader applications in understanding human cognition and behavior.