Yujie Wang ¹ CHENG-BANG CHEN ^1* Toshihiro Imamura ^2,3 Ignacio E. Tapia ⁴ Virend K. Somers ⁵ Phyllis C. Zee ⁶ Diane C. Lim ^7,8

• ¹ Department of Industrial and Systems Engineering, College of Engineering, University of Miami, Coral Gables, Florida, United States
• ² Division of Sleep Medicine, Department of Medicine, Perelman School of Medicine, University of Pennsylvania, Phialdelphia, Pennsylvania, United States
• ³ Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
• ⁴ Division of Pediatric Pulmonology, Miller School of Medicine, University of Miami, Miami, United States
• ⁵ Department of Cadiovascular Medicine, Mayo Clinic, Rochester, Michigan, United States
• ⁶ Center for Circadian and Sleep Medicine, Department of Neurology, Feinberg School of Medicine, Northwestern University, Chicago, United States
• ⁷ Miami VA Healthcare System, Veterans Health Administration, United States Department of Veterans Affairs, Miami, Florida, United States
• ⁸ Department of Medicine, Miller School of Medicine, University of Miami, Miami, United States

Objective: Recognizing emotions from electroencephalography (EEG) signals is a challenging task due to the complex, nonlinear, and nonstationary characteristics of brain activity. Traditional methods often fail to capture these subtle dynamics, while deep learning approaches lack explainability. In this research, we introduce a novel three-phase methodology integrating manifold embedding, multilevel heterogeneous recurrence analysis (MHRA), and ensemble learning to address these limitations in EEG-based emotion recognition. Approach: The proposed methodology was evaluated using the SJTU-SEED IV database. We first applied uniform manifold approximation and projection (UMAP) for manifold embedding of the 62-lead EEG signals into a lower-dimensional space. We then developed MHRA to characterize the complex recurrence dynamics of brain activity across multiple transition levels. Finally, we employed tree-based ensemble learning methods to classify four emotions (neutral, sad, fear, happy) based on the extracted MHRA features. Main results: Our approach achieved high performance, with an accuracy of 0.7885 and an AUC of 0.7552, outperforming existing methods on the same dataset.Additionally, our methodology provided the most consistent recognition performance across different emotions. Sensitivity analysis revealed specific MHRA metrics that were strongly associated with each emotion, offering valuable insights into the underlying neural dynamics.Significance: This study presents a novel framework for EEG-based emotion recognition that effectively captures the complex nonlinear and nonstationary dynamics of brain activity while maintaining explainability. The proposed methodology offers significant potential for advancing our understanding of emotional processing and developing more reliable emotion recognition systems with broad applications in healthcare and beyond.