Epilepsy is a long-term neurological condition marked by recurrent seizures, which result from abnormal electrical activity in the brain that disrupts its normal functioning. Traditional methods for detecting epilepsy through machine learning typically utilize discrete-time models, which inadequately represent the continuous dynamics of electroencephalogram (EEG) signals. To overcome this limitation, we introduce an innovative approach that employs Neural Ordinary Differential Equations (NODEs) to model EEG signals as continuous-time systems. This allows for effective management of irregular sampling and intricate temporal patterns. In contrast to conventional techniques, such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), which necessitate fixed-length inputs and often struggle with long-term dependencies, our framework incorporates: (1) a NODE block to capture continuous-time EEG dynamics, (2) a feature extraction module tailored for seizure-specific patterns, and (3) an attention-based fusion mechanism to enhance interpretability in classification. When evaluated on three publicly accessible EEG datasets, including those from Boston Children’s Hospital and the Massachusetts Institute of Technology (CHB-MIT) and the Temple University Hospital (TUH) EEG Corpus, the model demonstrated an average accuracy of 98.2%, a sensitivity of 97.8%, a specificity of 98.3%, and an F1-score of 97.9%. Additionally, the inference latency was reduced by approximately 30% compared to standard CNN and Long Short-Term Memory (LSTM) architectures, making it well-suited for real-time applications. The method’s resilience to noise and its adaptability to irregular sampling enhance its potential for clinical use in real-time settings.

Email communication plays a crucial role in both personal and professional contexts; however, it is frequently compromised by the ongoing challenge of spam, which detracts from productivity and introduces considerable security risks. Current spam detection techniques often struggle to keep pace with the evolving tactics employed by spammers, resulting in user dissatisfaction and potential data breaches. To address this issue, we introduce the Divide and Conquer-Generative Adversarial Network Squeeze and Excitation-Based Framework (DaC-GANSAEBF), an innovative deep-learning model designed to identify spam emails. This framework incorporates cutting-edge technologies, such as Generative Adversarial Networks (GAN), Squeeze and Excitation (SAE) modules, and a newly formulated Light Dual Attention (LDA) mechanism, which effectively utilizes both global and local attention to discern intricate patterns within textual data. This approach significantly improves efficiency and accuracy by segmenting scanned email content into smaller, independently evaluated components. The model underwent training and validation using four publicly available benchmark datasets, achieving an impressive average accuracy of 98.87%, outperforming leading methods in the field. These findings underscore the resilience and scalability of DaC-GANSAEBF, positioning it as a viable solution for contemporary spam detection systems. The framework can be easily integrated into existing technologies to enhance user security and reduce the risks associated with spam.