Neuromarketing is an emerging field that combines neuroscience and marketing
to understand the factors that influence consumer decisions better. The study
proposes a method to understand consumers' positive and negative reactions to
advertisements (ads) and products by analysing electroencephalogram (EEG)
signals. These signals are recorded using a low-cost single electrode headset
from volunteers belonging to the ages 18-22. A detailed subject dependent (SD)
and subject independent (SI) analysis was performed employing machine learning
methods like Naive Bayes (NB), Support Vector Machine (SVM), k-nearest
neighbour and Decision Tree and the proposed deep learning (DL) model. SVM and
NB yielded an accuracy (Acc.) of 0.63 for the SD analysis. In SI analysis, SVM
performed better for the advertisement, product and gender-based analysis.
Furthermore, the performance of the DL model was on par with that of SVM,
especially, in product and ads-based analysis.

本研究提出了一种理解消费者对广告和产品的正面和负面反应的方法，通过从 18-22 岁志愿者使用低成本单电极头戴设备记录的脑电图（EEG）信号进行分析。使用朴素贝叶斯（NB）、支持向量机（SVM）、k 最近邻和决策树等机器学习方法进行了详细的基于受试者和基于人口统计学的分析，同时提出了一种深度学习模型。结果表明，在基于广告、产品和性别的组别分析中，SVM 表现更好，而 DL 模型的性能与 SVM 相当，尤其在产品和广告分析方面表现得很好。