This study proposes a novel multimodal fatigue classification framework that integrates differential entropy (DE) features from electroencephalogram (EEG) signals and heart rate variability (HRV) features from electrocardiogram (ECG) signals. Correlation coefficient matrices are calculated between DE and HRV features using Laplacian eigenvalues and singular value decomposition (SVD). Two different electrode configurations were tested: 64-channel and 17-channel. Each configuration features one channel for the ECG signal, while the remaining are for the EEG signals. Results demonstrated that integrating both EEG and ECG features outperforms using a single modality. Changes in brain activity are measured using EEG signals, and changes in the state of the autonomic nervous system are measured using ECG signals. Thus, this model captures CNS and autonomic nervous system symptoms of fatigue.