Electroencephalogram signals are routinely used to monitor brain activity in response to stimuli. The primary goal of taste-related brain-computer interfaces is to detect taste sensations. The Taste-EEG dataset comprises EEG signals captured during taste stimulation, making it a fascinating arena for researching brain responses to gustatory stimuli. We suggest using the Discrete Wavelet Transform (DWT) as a feature extraction approach to solve this issue. We can extract temporal and spectral information from EEG data using DWT. This allows us to investigate and comprehend how the brain reacts to various flavors. The collected characteristics are then used to train and test basic neural network models. Neural networks have proven effective in learning intricate patterns and relationships within data, making them well-suited for accurately classifying taste-related brain activity. Our experimental results demonstrate that our approach accurately distinguishes between sour and salty tastes. This indicates that we can effectively capture discriminative information from EEG signals associated with taste perception by combining DWT-based feature extraction with neural network models. This research significantly advances taste-related brain-computer interfaces by providing a robust methodology for accurately classifying taste sensations based on EEG signals recorded during taste stimulation.