Emotions are fundamental to human cognition and behavior, significantly influencing mental well-being. There is a growing research interest in emotion analysis due to its diverse applications in domains such as mental health, human–computer interaction, and affective computing. This study proposes a deep forest (DF)-based emotion recognition system (ERS) that identifies emotions from electroencephalography (EEG) signals. The proposed system aims to recognize emotions and contribute to mental well-being by analyzing brain activity linked to various emotional states. Experiments were conducted on the Brain Wave Dataset (Dataset Availability Statement: Source link: https://www.kaggle.com/datasets/birdy654/eeg-brainwave-dataset-feeling-emotions ), which features recordings of electrical signals captured by scalp electrodes under three emotional states: positive, neutral, and negative. The results demonstrate that the proposed multi-class classifier outperforms conventional machine learning models, showcasing promising potential for accurate emotion recognition from EEG signals for mental well-being.