Effective Model Research Articles

Effective Emotion Recognition by Learning Discriminative Graph Topologies in EEG Brain Networks.

Multichannel electroencephalogram (EEG) is an array signal that represents brain neural networks and can be applied to characterize information propagation patterns for different emotional states. To reveal these inherent spatial graph features and increase the stability of emotion recognition, we propose an effective emotion recognition model that performs multicategory emotion recognition with multiple emotion-related spatial network topology patterns (MESNPs) by learning discriminative graph topologies in EEG brain networks. To evaluate the performance of our proposed MESNP model, we conducted single-subject and multisubject four-class classification experiments on two public datasets, MAHNOB-HCI and DEAP. Compared with existing feature extraction methods, the MESNP model significantly enhances the multiclass emotional classification performance in the single-subject and multisubject conditions. To evaluate the online version of the proposed MESNP model, we designed an online emotion monitoring system. We recruited 14 participants to conduct the online emotion decoding experiments. The average online experimental accuracy of the 14 participants was 84.56%, indicating that our model can be applied in affective brain-computer interface (aBCI) systems. The offline and online experimental results demonstrate that the proposed MESNP model effectively captures discriminative graph topology patterns and significantly improves emotion classification performance. Moreover, the proposed MESNP model provides a new scheme for extracting features from strongly coupled array signals.

Identification and validation of plasma AGRN as a novel diagnostic biomarker of hepatitis B Virus-related chronic hepatitis and liver fibrosis/cirrhosis.

The aim of this study was to find novel biomarkers and develop a non-invasive, effective diagnostic model for hepatitis B Virus-related chronic hepatitis and liver fibrosis/cirrhosis. Quantitative real-time polymerase chain reaction (qRT-PCR) was utilized to assess the expression of differentially expressed genes (AGRN, JAG1, CCL5, ID3, CCND1, and CAPN2) in peripheral blood mononuclear cells (PBMCs) from healthy subjects, chronic hepatitis B (CHB), and liver fibrosis/cirrhosis (LF/LC) patients. The molecular mechanisms underlying AGRN-regulated CHB were further explored and verified in LX2 cells, in which small interfering RNA (siRNA) was used to block AGRN gene expression. Finally, enzyme-linked Immunosorbent Assay (ELISA) was used to measure AGRN protein expression in 100 healthy volunteers, 100 CHB patients, and 100 LF/LC patients, and the efficacy of the diagnostic model was assessed by the Area Under the Curve (AUC). AGRN mRNA displayed a steady rise in the PBMCs of normal, CHB, and LF/LC patients. Besides, AGRN expression was markedly elevated in activated LX2 cells, whereas the expression of COL1 and α-SMA decreased when AGRN was inhibited using siRNA. In addition, downregulation of AGRN can reduce the gene expression of β-catenin and c-MYC while upregulating the expression of GSK-3β. Furthermore, PLT and AGRN were used to develop a non-invasive diagnostic model (PA). To identify CHB patients from healthy subjects, the AUC of the PA model was 0.951, with a sensitivity of 87.0% and a specificity of 91.0%. The AUC of the PA model was 0.922 with a sensitivity of 82.0% and a specificity of 90.0% when differentiating between LF/LC and CHB patients. The current study indicated that AGRN could be a potential plasma biomarker and the established PA model could improve the diagnostic accuracy for HBV-related liver diseases.

Accuracy of death risk prediction models for acute coronary syndrome patients: a systematic review and meta-analysis.

This study systematically evaluates the accuracy of several death risk prediction models for patients with acute coronary syndrome (ACS) through evidence-based methods. We identify the most accurate and effective ACS death risk prediction model and provide an evidence-based basis for clinical healthcare personnel to evaluate their choice of death risk prediction model for ACS patients. An evidence-based approach was used to study the current death risk prediction model for ACS. First, a literature search was carried out using computer-based and manual searching. The literature databases searched include Cochrane Library, MEDLINE, EMBASE, PubMed, Web of Science, WanFang Data, CNKI, VPCS, and SinoMed. The search period was limited to 2009 to 2022. Screening, quality evaluation and data extraction were carried out for the included articles. The PROBAST was used to conduct a migration risk assessment. RevMan 5.3 and Meta-DiSc 1.4 were used in combination to determine the model effect sizes. A descriptive analysis was conducted for the data that could not be meta-analyzed. A total of 8277 articles were initially included in this study. After screening, 25 articles were finally included, involving 11 different risk prediction models. A total of 306,390 patients with ACS were included of which 158,080 (51.6%) were male and 147,793 (48.4%) were female. The patients stemmed from 11 different countries (e.g., China, the USA, Spain, the UK, etc.). The total number of deaths was 23,601. The sensitivity of the GRACE risk prediction model was 0.78, with a specificity of 0.76 and an AUC value of 0.86. The sensitivity of the CAMI risk prediction model was 0.78, with a specificity of 0.70 and an AUC value of 0.85. The sensitivity of the TIMI risk prediction model was 0.51, with a specificity of 0.81, and an AUC value of 0.64. The sensitivity of the REMS risk prediction model was 0.78, with a specificity of 0.46 and an AUC value of 0.41. Eight different risk prediction models (EPICOR, CRUSADE, SAMI, GWTG, LNS, SYNTAX II, APACHE II) that could not be combined with the effect size were also included, with sensitivities ranging from 0.77-0.95, specificities ranging from 0.22-0.99, and AUC values ranging from 0.71-0.92. The GRACE and CAMI risk prediction models demonstrate good accuracy for evaluating the death risk of ACS patients. The accuracy of the TIMI risk prediction model is similar to that of the REMS risk prediction model. The APACHE II, SYNTAX II, EPICOR, and CAMI risk prediction models also show good accuracy for estimating the risk of death in ACS patients, although further validation is needed due to limited evidence. For improved predictive accuracy and to help advance medical interventions, the author recommends that clinical medical staff use the GRACE model to predict the death risk of ACS patients.

CRE-TSCAE: A Novel Classification Model Based on Stacked Convolutional Autoencoder for Dual-Target RSVP-BCI Tasks.

The Rapid Serial Visual Presentation (RSVP) paradigm facilitates target identification in a rapid picture stream, which is applied extensively in military target surveillance and police monitoring. Most researchers concentrate on the single target RSVP-BCI whereas the study of dual-target is scarcely conducted, limiting RSVP application considerably. This paper proposed a novel classification model named Common Representation Extraction-Targeted Stacked Convolutional Autoencoder (CRE-TSCAE) to detect two targets with one nontarget in RSVP tasks. CRE generated a common representation for each target class to reduce variability from different trials of the same class and distinguish the difference between two targets better. TSCAE aimed to control uncertainty in the training process while requiring less target training data. The model learned a compact and discriminative feature through the training from several learning tasks so as to distinguish each class effectively. It was validated on the World Robot Contest 2021 and 2022 ERP datasets. Experimental results showed that CRE-TSCAE outperformed the state-of-the-art RSVP decoding algorithms and the Average ACC was 71.25%, improving 6.5% at least over the rest. It demonstrated that CRE-TSCAE showed a strong ability to extract discriminative latent features in detecting the differences among two targets with nontarget, which guaranteed increased classification accuracy. CRE-TSCAE provided an innovative and effective classification model for dual-target RSVP-BCI tasks and some insights into the neurophysiological distinction between different targets.

Non-Autoregressive Line-Level Code Completion

Software developers frequently use code completion tools to accelerate software development by suggesting the following code elements. Researchers usually employ AutoRegressive (AR) decoders to complete code sequences in a left-to-right, token-by-token fashion. To improve the accuracy and efficiency of code completion, we argue that tokens within a code statement have the potential to be predicted concurrently. In this article, we first conduct an empirical study to analyze the dependency among the target tokens in line-level code completion. The results suggest that it is potentially practical to generate all statement tokens in parallel. To this end, we introduce SANAR, a simple and effective syntax-aware non-autoregressive model for line-level code completion. To further improve the quality of the generated code, we propose an adaptive and syntax-aware sampling strategy to boost the model’s performance. The experimental results obtained from two widely used datasets indicate that our model outperforms state-of-the-art code completion approaches of similar model size by a considerable margin, and is faster than these models with up to 9× speed-up. Moreover, the extensive results additionally demonstrate that the enhancements achieved by SANAR become even more pronounced with larger model sizes, highlighting their significance.

Cross-sectoral role and community empowerment in controlling schistosomiasis in Indonesia

The role of cross-sectors in synergy with community empowerment in controlling schistosomiasis is a very effective model. Until mid-2017, the incidence rate of disease in humans in 28 endemic villages still ranged 0.00-2.15%. This research aims to evaluate schistosomiasis control activities that have been carried out across sectors based on a road map for schistosomiasis eradication and community empowerment in controlling schistosomiasis. The study was mixed method. Data were obtained by in-depth interview with key informants and evaluation of the results of cross-sectoral and community activities. The research was carried out in an endemic area for schistosomiasis on the Bada Plateau, Indonesia, from January to November 2019. In conclusion, the potential of cross-sector collaboration in controlling schistosomiasis has not been fully realized due to budget constraints. Community empowerment is carried out to eliminate snail foci, increase the scope of feces collection and change community behavior for the better in order to avoid infection and always try to clean up snail foci.

DAGCN: Dynamic and Adaptive Graph Convolutional Network for Salient Object Detection.

Deep-learning-based salient object detection (SOD) has achieved significant success in recent years. The SOD focuses on the context modeling of the scene information, and how to effectively model the context relationship in the scene is the key. However, it is difficult to build an effective context structure and model it. In this article, we propose a novel SOD method called dynamic and adaptive graph convolutional network (DAGCN) that is composed of two parts, adaptive neighborhood-wise graph convolutional network (AnwGCN) and spatially restricted K-nearest neighbors (SRKNN). The AnwGCN is novel adaptive neighborhood-wise graph convolution, which is used to model and analyze the saliency context. The SRKNN constructs the topological relationship of the saliency context by measuring the non-Euclidean spatial distance within a limited range. The proposed method constructs the context relationship as a topological graph by measuring the distance of the features in the non-Euclidean space, and conducts comparative modeling of context information through AnwGCN. The model has the ability to learn the metrics from features and can adapt to the hidden space distribution of the data. The description of the feature relationship is more accurate. Through the convolutional kernel adapted to the neighborhood, the model obtains the structure learning ability. Therefore, the graph convolution process can adapt to different graph data. Experimental results demonstrate that our solution achieves satisfactory performance on six widely used datasets and can also effectively detect camouflaged objects. Our code will be available at: https://github.com/CSIM-LUT/DAGCN.git.

A Global and Local Surrogate-Assisted Genetic Programming Approach to Image Classification

Genetic programming (GP) has achieved promising performance in image classification. However, GP-based methods usually require a long computation time for fitness evaluations, posing a challenge to real-world applications. Surrogate models can be efficiently computable approximations of expensive fitness evaluations. However, most existing surrogate methods are designed for evolutionary computation techniques with a vector-based representation consisting of numerical values, thus cannot be directly used for GP with a tree-based representation consisting of functions/operators. The variable sizes of GP trees further increase the difficulty of building the surrogate model for fitness approximations. To address these limitations, we propose a new surrogate-assisted GP approach including global and local surrogate models, which can accelerate the evolutionary learning process and achieve competitive classification performance simultaneously. The global surrogate model can assist GP in exploring the entire search space, while the local surrogate model can speed up convergence and further improve performance. Furthermore, a new surrogate training set is constructed to assist in establishing the relationship between the GP tree and its fitness, and effective surrogate models can be built accordingly. Experimental results on ten datasets of varying difficulty show that the new approach significantly reduces the computational cost of the GP-based method without sacrificing the classification accuracy. The comparisons with other state-of-the-art methods also demonstrate the effectiveness of the new approach. Further analysis reveals the significance of the global and local surrogates and the new surrogate training set on improving or maintaining the performance of the proposed approach while reducing the computational cost.

Fault Detection of Urban Wastewater Treatment Process Based on Combination of Deep Information and Transformer Network.

As one of the hot issues of concerns during modern social development, the wastewater treatment process is acknowledged to be a process with complex biochemical reactions and susceptible to an external environment, featuring strong nonlinear and time correlation characteristics, which are difficult for traditional mechanism-based models to tackle. For many classical data-driven fault detection methods, a complete retraining process is necessary to monitor every new fault, and most of the current neural network-based strategies rarely achieve satisfactory monitoring accuracy or robustness either. Giving full consideration to the aforementioned problems, this article takes advantage of position encoding, residual connection, and multihead attention mechanism embedded in the Transformer structure to establish an effective and efficient wastewater treatment process fault detection model, where offline modeling and online monitoring are performed successively to achieve accurate detection of the faults. In the experimental part, the advantages of the proposed method are strongly verified through the simulation monitoring of 27 faults on the benchmark simulation model 1 (BSM1), where the false alarm rate (FAR) and miss alarm rate (MAR) of the established method are proved to be significantly lower than those of the compared state-of-the-art methods.

Uncertainty-Aware Aggregation for Federated Open Set Domain Adaptation.

Open set domain adaptation (OSDA) methods have been proposed to leverage the difference between the source and target domains, as well as to recognize the known and unknown classes in the target domain. Such methods typically require the entire source and target data simultaneously to train the target model. However, in real scenarios, data are distributed and stored in various clients. They cannot be exchanged among clients because of privacy protection. Federated learning (FL) is a decentralized approach for training an effective global model with the training data distributed among the clients. Despite its potential in addressing the privacy concerns of data sharing, FL methods for OSDA that can handle unknown classes is not yet available. To tackle this problem, we have developed a novel federated OSDA (FOSDA) algorithm. More specifically, FOSDA adopts an uncertainty-aware mechanism to generate a global model from all client models. It reduces the uncertainty of the federated aggregation by focusing on the contribution of source clients with high uncertainty while retaining those with high consistency. Moreover, a federated class-based weighted strategy is also implemented in FOSDA to maintain the category information of the source clients. We have conducted comprehensive experiments on three benchmark datasets to evaluate the performance of the proposed method, and the results demonstrate the effectiveness of FOSDA.

PDHL-MARCOS method based on cumulative prospect theory for MAGDM and its application in network security evaluation

The increasingly large and complex network brings convenience to our life and production, but it also brings potential dangers. Any minor vulnerability may lead to failure of application software configuration, software maintenance, user management and system management. Eventually, our network is interrupted, and even life and production data are leaked. Therefore, establishing a scientific and effective network security evaluation model is an effective way to supervise, manage and improve network security; It is also an effective measure to reduce the loss of life and production. Therefore, we first selected six factors closely related to network security to form a complete network security evaluation index system. Then, the utility function in the traditional measurement of alternatives and ranking according to compromise solution (MARCOS) method is improved by using the value function in cumulative prospect theory (CPT), and a PDHL-MARCOS evaluation model that can reflect the decision-makers’ (DMs’) attitudes towards gains and losses is obtained in the probabilistic double hierarchy linguistic term set (PDHLTS). In addition, the relative entropy (RE) method is also used to obtain expert weights. The combination weights of attributes are obtained using the correlation coefficient and the standard deviation (CCSD) method and the weight function in CPT.

CMCEE: A joint learning framework for cascade decoding with multi-feature fusion and conditional enhancement for overlapping event extraction

Event extraction (EE) is an important natural language processing task. With the passage of time, many powerful and effective models for event extraction tasks have been developed. However, there has been limited research on complex overlapping event extraction. Therefore, we propose a new cascade decoding model: A Joint Learning Framework for Cascade Decoding with Multi-Feature Fusion and Conditional Enhancement for Overlapping Event Extraction. 1) In this model, we introduce a cascade decoding mechanism with multi-feature fusion to better capture the interaction between decoding layers. 2) Additionally, we introduce an enhanced conditional layer normalization (ECLN) mechanism to enhance the interaction between subtasks. Simultaneously, the use of a cascade decoding model effectively addresses the problem of overlapping events. The model successively performs three subtasks, type detection, trigger word extraction and argument extraction. All three subtasks learned together in a framework, and a new conditional normalization mechanism is used to capture dependencies among these subtasks. The experiments are conducted using the overlapping event benchmark, FewFC dataset. The experimental evaluation demonstrates that our model achieves a higher F1 score on the overlapping event extraction task compared to the original overlapping event extraction model.

Silver-graphene oxide nanocomposite doping chitosan/PVA membrane for arsenic (III) elimination from aqueous solution

Heavy metals and pathogens from contaminated water sources may undoubtedly be removed by creating an efficient bio-adsorbent based on functional spots. Thus, the goal of this work was to produce chitosan (Ch)-polyvinyl alcohol (PVA) biofilm decorated with graphene oxide (GO) sheets doped with silver nanoparticles (AgNPs). The nanostructure of prepared GO/Ag nanosheets is examined by transmission electron microscope (TEM). The fabricated film (GO/Ag Ch-PVA) is compared by the control films (Ch, PVA and Ch-PVA). Fourier transform infrared (FTIR) spectroscopy, X-ray diffraction (XRD), and tensile strength are used to study the films’ structure. Also, the antimicrobial activity was assessed for the films. After doping the polymer matrix with GO/Ag, it was discovered that the tensile strength increased to about 46.18 MPa. Moreover, adsorption experiment for arsenic As (III) ions is explored by the prepared film at different operating conditions. The obtained results validated the enhanced adsorption ability of the GO/Ag Ch-PVA film towards As (III) with the highest adsorption capacity of 54.3 mg g−1 obtained from the isotherm model of Langmuir. Moreover, kinetic mathematical models for the adsorption effectiveness of GO/Ag Ch-PVA film are assessed. The results gathered demonstrated that GO/Ag Ch-PVA film is a potentially useful material for eliminating As (III) and microbial strains from essential water resources.

Fast Graph Generation via Spectral Diffusion.

Generating graph-structured data is a challenging problem, which requires learning the underlying distribution of graphs. Various models such as graph VAE, graph GANs, and graph diffusion models have been proposed to generate meaningful and reliable graphs, among which the diffusion models have achieved state-of-the-art performance. In this paper, we argue that running full-rank diffusion SDEs on the whole graph adjacency matrix space hinders diffusion models from learning graph topology generation, and hence significantly deteriorates the quality of generated graph data. To address this limitation, we propose an efficient yet effective Graph Spectral Diffusion Model (GSDM), which is driven by low-rank diffusion SDEs on the graph spectrum space. Our spectral diffusion model is further proven to enjoy a substantially stronger theoretical guarantee than standard diffusion models. Extensive experiments across various datasets demonstrate that our proposed GSDM turns out to be the SOTA model, by exhibiting both significantly higher generation quality and much less computational consumption than the baselines.

Collaborative Knowledge Distillation via Multiknowledge Transfer.

Knowledge distillation (KD), as an efficient and effective model compression technique, has received considerable attention in deep learning. The key to its success is about transferring knowledge from a large teacher network to a small student network. However, most existing KD methods consider only one type of knowledge learned from either instance features or relations via a specific distillation strategy, failing to explore the idea of transferring different types of knowledge with different distillation strategies. Moreover, the widely used offline distillation also suffers from a limited learning capacity due to the fixed large-to-small teacher-student architecture. In this article, we devise a collaborative KD via multiknowledge transfer (CKD-MKT) that prompts both self-learning and collaborative learning in a unified framework. Specifically, CKD-MKT utilizes a multiple knowledge transfer framework that assembles self and online distillation strategies to effectively: 1) fuse different kinds of knowledge, which allows multiple students to learn knowledge from both individual instances and instance relations, and 2) guide each other by learning from themselves using collaborative and self-learning. Experiments and ablation studies on six image datasets demonstrate that the proposed CKD-MKT significantly outperforms recent state-of-the-art methods for KD.

Data Augmentation for Sample Efficient and Robust Document Ranking

Contextual ranking models have delivered impressive performance improvements over classical models in the document ranking task. However, these highly over-parameterized models tend to be data-hungry and require large amounts of data even for fine-tuning. In this article, we propose data-augmentation methods for effective and robust ranking performance. One of the key benefits of using data augmentation is in achieving sample efficiency or learning effectively when we have only a small amount of training data. We propose supervised and unsupervised data augmentation schemes by creating training data using parts of the relevant documents in the query-document pairs. We then adapt a family of contrastive losses for the document ranking task that can exploit the augmented data to learn an effective ranking model. Our extensive experiments on subsets of the MS MARCO and TREC-DL test sets show that data augmentation, along with the ranking-adapted contrastive losses, results in performance improvements under most dataset sizes. Apart from sample efficiency, we conclusively show that data augmentation results in robust models when transferred to out-of-domain benchmarks. Our performance improvements in in-domain and more prominently in out-of-domain benchmarks show that augmentation regularizes the ranking model and improves its robustness and generalization capability.

Overcoming Barriers to Collaborative Team Practice: A System Approach.

All healthcare systems are challenged to provide sufficient access to appropriate care for the individuals and communities they serve. Among the commonly discussed interventions in an era of clinical shortages is the establishment of team-based care, where team members can practice at the top of their license. This solution ensures that talent and ability are amplified. However, this vision has been thwarted by several barriers. Recognizing that complex causes prevent team-based care, CommonSpirit Health leadership has embarked on a multifaceted implementation of tactics to mitigate the barriers. The strategy addresses issues ranging from regulations, laws, and payment practices to a lack of knowledge and understanding between professions. A combination of solutions, rather than discrete tactics, holds the key to the system's strategy. The complement of actions put in place at CommonSpirit Health has included new executive roles, dyad leadership models, shared multidisciplinary education, targeted advocacy, best practice playbooks, and the flexibility to individualize local models as part of the journey to a more effective and efficient model of care.

CNN-CatBoost ensemble deep learning model for enhanced disease detection and classification of kidney disease

An efficient deep-learning prediction model for identifying chronic kidney disease (CKD) from exhaled breath is presented in this paper. The concentration of urea will be higher in CKD patients. Salivary urease breaks down the stored urea into ammonia, which is then excreted through breath. Thus, by monitoring the breath ammonia content, it is possible to identify the presence of high urea levels in the body. In this work, a novel sensing module is developed and applied to measure and assess the amount of ammonia in exhaled breath. Moreover, an effective deep learning prediction model that combines the CatBoost algorithm and convolutional neural network (CNN) is used to automate the prediction of disease. The proposed model, which combines the benefits of gradient-boosting and CNN, attained an exceptional accuracy of 98.37%. Experiments are conducted to evaluate the proposed model using real-time data and to assess how well it performs in comparison with existing deep learning methods. Our study's findings demonstrate that kidney disease can be accurately and noninvasively diagnosed using the proposed approach.

Improving student higher order thinking skills using SynecticHOTS-oriented learning model

The achievement of students’ learning outcomes in physics lessons has a low tendency, this is due to various factors, one of which is higher-order thinking skills (HOTS) based learning that has not been fully implemented in the learning process and assessment. This study aimed to determine the differences in the Synectic-HOTS learning model in physics compared to conventional learning models. This study used a quasi-experimental study with a pre-test–post-test nonequivalent control group design. The instrument used in this study is a two-tier test question that contains 20 items of HOTS aspect. The prerequisite test showed that the data was not homogeneous, while the results of the data homogeneity test were stated to be homogeneous. Therefore, a non-parametric test was carried out using the Mann-Whitney U-test. Based on the results of the data analysis, the Asymp. Sig. (2-tailed) for the pre-test was 0.111 and the post-test was 0.001 (p=.000<05). The effect size calculation shows that the Synectic learning model is effective in the medium category. These results indicate that the application of the SynecticHOTS model significantly improves HOTS when compared to conventional learning models. This work can open up insights to teachers about alternative, creative, and effective learning models to improve higher-order thinking skills and student learning outcomes.

Hierarchical and dynamic graph attention network for drug-disease association prediction.

In the realm of biomedicine, the prediction of associations between drugs and diseases holds significant importance. Yet, conventional wet lab experiments often fall short of meeting the stringent demands for prediction accuracy and efficiency. Many prior studies have predominantly focused on drug and disease similarities to predict drug-disease associations, but overlooking the crucial interactions between drugs and diseases that are essential for enhancing prediction accuracy. Hence, in this paper, a resilient and effective model named Hierarchical and Dynamic Graph Attention Network (HDGAT) has been proposed to predict drug-disease associations. Firstly, it establishes a heterogeneous graph by leveraging the interplay of drug and disease similarities and associations. Subsequently, it harnesses the capabilities of graph convolutional networks and bidirectional long short-term memory networks (Bi-LSTM) to aggregate node-level information within the heterogeneous graph comprehensively. Furthermore, it incorporates a hierarchical attention mechanism between convolutional layers and a dynamic attention mechanism between nodes to learn embeddings for drugs and diseases. The hierarchical attention mechanism assigns varying weights to embeddings learned from different convolutional layers, and the dynamic attention mechanism efficiently prioritizes inter-node information by allocating each node with varying rankings of attention coefficients for neighbour nodes. Moreover, it employs residual connections to alleviate the over-smoothing issue in graph convolution operations. The latent drug-disease associations are quantified through the fusion of these embeddings ultimately. By conducting 5-fold cross-validation, HDGAT's performance surpasses the performance of existing state-of-the-art models across various evaluation metrics, which substantiates the exceptional efficacy of HDGAT in predicting drug-disease associations.