Understanding and analyzing human emotions is a multifaceted yet critical aspect of applications, including artificial intelligence, healthcare, and human–computer interaction. Electroencephalogram (EEG) signals have become crucial for comprehending emotional states, offering an unobstructed view of the brain processes associated with distinct emotions. Nonetheless, the complex nature of EEG data requires the application of advanced algorithms to infer meaningful patterns and improve the precision of emotion recognition. The use of EEG data for emotion detection has emerged as a potentially productive technique for understanding and interpreting human affective states. Using machine learning algorithms, this research study proposes a better method for emotion identification based on EEG data. The Random Forest (RF) technique is employed as a robust classification model, and feature extraction is performed through the utilize of the Fast Fourier Transform (FFT). The FFT provides a comprehensive portrayal of neural activity that is fundamentally linked to diverse emotional states by analyzing the frequency components of EEG signals. The Random Forest algorithm’s ensemble learning capabilities and adaptability are then utilized to classify these data into discrete emotional categories. The experimental results using a dataset of varied emotive stimuli support the efficacy of the proposed approach. The combination of FFT and RF algorithms not only enhances the precision of emotion recognition but also progresses the model’s ability to generalize results across a variety of experimental conditions and individuals.