Interictal Epileptiform Discharge is essential for identifying epilepsy. However, the unpredictable and non-stationary nature of electroencephalogram (EEG) patterns poses considerable challenges for reliable identification. Manual interpretation of EEG is subjective and time-consuming. With advancements in machine learning and deep learning, computer-aided approaches for automated IED detection have been rapidly developed. The state-of-the-art convolutional neural network (CNN)-based methods have shown promising results but struggle to capture long-term dependencies in time-series data. In contrast, Transformer excels at modeling sequential information through self-attention mechanisms, overcoming the CNN limitations. This study proposes an IED Detector (IEDD) that integrates convolutional layers and a Transformer to detect IEDs. The IEDD initially employs convolutional layers to extract local features of IEDs, followed by a Transformer to model long-term dependencies. To further extract spatial features, EEG data are represented as a three-dimensional tensor with embedded channel topology, where a CNN captures spatial features at each sampling point and a Long Short-Term Memory (LSTM) network models their temporal evolution. Additionally, due to the scarcity of IED data, a novel Transformer-based Generative Adversarial Network (GAN) is developed to augment the IED dataset. Experimental results show the proposed approach achieves an average accuracy of 96.11% on the augmented Dataset 1 and 95.25% on Dataset 2 for binary classification, with an average sensitivity of 87.26% and precision of 89.96% for multi-label classification. These findings provide valuable insights into advancing deep learning and Transformer-based approaches for automated IED detection.

Multimodal Large Language Models (MLLMs) have achieved strong performance across vision-language tasks, but suffer from significant computational overhead due to the quadratic growth of attention computations with the number of multimodal tokens.Though efforts have been made to prune tokens in MLLMs, they lack a fundamental understanding of how MLLMs process and fuse multimodal information.Through systematic analysis, we uncover a three-stage cross-modal interaction process: (1) Shallow layers recognize task intent, with visual tokens acting as passive attention sinks; (2) Cross-modal fusion occurs abruptly in middle layers, driven by a few critical visual tokens; (3) Deep layers discard vision tokens, focusing solely on linguistic refinement.Based on these findings, we propose VisiPruner, a training-free pruning framework that reduces up to 99% of visionrelated attention computations and 53.9% of FLOPs on LLaVA-v1.5 7B.It significantly outperforms existing token pruning methods and generalizes across diverse MLLMs.Beyond pruning, our insights further provide actionable guidelines for training efficient MLLMs by aligning model architecture with its intrinsic layer-wise processing dynamics.

Multimodal representation, as an application framework of social computing, enables researchers to utilize multimodal data for more accurate predictive analysis of mortality rates. In lengthy clinical texts, there are numerous neutral words that do not directly reflect the patient's condition. However, the prevalence of frequently occurring neutral words diminishes the weights of key terms that are directly relevant to the patient's condition, resulting in an imbalance in the allocation of text feature weights. To address this issue, we propose multimodal mortality prediction-multilevel semantic hypergraph (MMP-MSH), a medical multimodal model based on a multilevel semantic hypergraph. Specifically, we approach clinical text in two ways. First, we employ a CNN to extract textual features directly. Second, the text processed into hypergraph is subjected to multilevel GCN to obtain global hypergraph information, which is then introduced into the model training process and combined with the features of each batch of clinical texts. We conducted experiments on the MIMIC-III dataset to evaluate the effectiveness of MMP-MSH in predicting mortality rates.

In clinical acupuncture practice, needle twirling (NT) and needle retention (NR) are strategically combined to achieve different therapeutic effects, highlighting the importance of distinguishing between different acupuncture states. Scalp EEG has been proven significantly relevant to brain activity and acupuncture stimulation. In this work, we designed an acupuncture paradigm to collect scalp EEG to study the differences in EEG changes during different acupuncture states. Since deep learning (DL) has been increasingly used in EEG analysis, we propose the Acupuncture Transformer Detector (ATD), a model based on Convolutional Neural Networks (CNN) and Transformer technology. ATD encapsulates the local and global features of EEG under the acupuncture states of Zusanli acupoint (ST-36) in an end-to-end classification framework. The experiment results from 28 healthy participants show that the proposed model can efficiently classify the EEG in different states, with an accuracy of $85.47\pm 0.73\%$. In this study, time-frequency analysis revealed that power changes were mainly confined to the delta frequency band under different acupuncture states. Brain topography revealed that ST-36 was activated primarily on the left frontal and parieto-occipital areas. This method provides new ideas for automatic recognition of acupuncture status from the perspective of DL, offering new solutions for standardizing acupuncture procedures.

High-quality scalp EEG datasets are extremely valuable for motor imagery (MI) analysis. However, due to electrode size and montage, different datasets inevitably experience channel information loss, posing a significant challenge for MI decoding. A 2D representation that focuses on the time domain may loss the spatial information in EEG. In contrast, a 3D representation based on topography may suffer from channel loss and introduce noise through different padding methods. In this paper, we propose a framework called Reference Electrode Standardization Interpolation Network (REI-Net). Through an interpolation of 3D representation, REI-Net retains the temporal information in 2D scalp EEG while improving the spatial resolution within a certain montage. Additionally, to overcome the data variability caused by individual differences, transfer learning is employed to enhance the decoding robustness. Our approach achieves promising performance on two widely-recognized MI datasets, with an accuracy of 77.99% on BCI-C IV-2a and an accuracy of 63.94% on Kaya2018. The proposed algorithm outperforms the SOTAs leading to more accurate and robust results.

Dropout prediction is an important task due to the high attrition rate commonly found on the massive open online courses (MOOC) platforms. Researchers usually use neural networks to establish various prediction models based on the behavioral features of student data. However, the existing methods ignore the periodic feature of learning behaviors and the influence of learning time distribution information on the prediction results, resulting in the potential association relationship between the input data is not learned by the model. Thus, after in-depth analysis of MOOC learners' behavior data, this paper proposes the concept of periodic feature, and found that different learning time has different effects on the prediction results. Based on the gained insights, we propose a hybrid neural network model (CGDC-LSTM) to model and to predict users' dropout behavior. CGDC-LSTM utilizes Convolutional Neural Network (CNN) to maintain the local correlation of students' behavior, and uses a module combining Group Convolution and Dilated Causal Convolution to fit the periodic feature of students, and combines Long Short-Term Memory Network (LSTM) into the model to extract the learning time distribution information to capture the influence of different learning periods on the results. Experimental results on the KDD Cup 2015 dataset demonstrate that the proposed model shows better prediction performance compared to baseline methods.

Kernel selection is fundamental to the generalization performance of kernel-based learning algorithms. Approximate kernel selection is an efficient kernel selection approach that exploits the convergence property of the kernel selection criteria and the computational virtue of kernel matrix approximation. The convergence property is measured by the notion of approximate consistency. For the existing Nyström approximations, whose sampling distributions are independent of the specific learning task at hand, it is difficult to establish the strong approximate consistency. They mainly focus on the quality of the low-rank matrix approximation, rather than the performance of the kernel selection criterion used in conjunction with the approximate matrix. In this paper, we propose a novel Nyström approximate kernel selection algorithm by customizing a criterion-driven adaptive sampling distribution for the Nyström approximation, which adaptively reduces the error between the approximate and accurate criteria. We theoretically derive the strong approximate consistency of the proposed Nyström approximate kernel selection algorithm. Finally, we empirically evaluate the approximate consistency of our algorithm as compared to state-of-the-art methods.

Dynamic holographic three-dimensional (3-D) display, which features reconstructing 3-D images with full depth cues has great potential in various fields such as medical, and military industries obtained a broad attention in the last decades. Combing parallel computation techniques, hologram synthesis algorithms are capable of calculating diffraction and interference pattern dynamically or even in real-time manner. In addition, the development of various types of optical modulators including liquid crystal (LC)-based, opto-acoustic-typed, digital micromirror devices (DMDs), and opto-opto materials made it possible that device can load and display such holographic patterns dynamically. In accordance with the progress of algorithms and modulators, we can expect that dynamic holographic 3-D display will be able to reconstruct 3-D images with real-time refresh rate, full color, wide viewing angle, and large image size, and the system will be commercialized in a near future.