CONSUMER PREFERENCE DETECTION IN NEUROMARKETING USING EEG

Abstract

Electroencephalography (EEG) has made several strides in neuroscience. It is generally used to monitor, diagnose, and identify several neurological conditions. The advantage of EEG is that it can record brain waves with high resolution with an extremely ergonomic setup. Hence, it has become a favoured choice for multiple applications like detection of dementia, epilepsy, classification of motor imagery, neuromarketing, measuring cognitive attention, etc. This dissertation focuses on the application of neuromarketing, which is a fusion of neuroscience and marketing. Debatably unethical, neuromarketing has quickly gained traction in industry after several experiments found success. An example is Frito-Lay, who used neuroimaging to entirely re-evaluate their approach to marketing. Customers were shown products with different packaging and colours, and their responses were recorded as positive, negative, or neutral. It was found that shiny packaging was not preferred, and matte was. Frito-Lay went on to scrap shiny packaging and adopted the matte look. The publicly available dataset recorded by Yadava et al [1] has been used. Independent component analysis (ICA), empirical mode decomposition (EMD) and logistic regression were subsequently applied on the raw data, and the best f1 score is reported as 89.41%, which is superior to the method used by Yadava et al [1], who achieved an accuracy of 70%.