Hierarchical Interactions Research Articles

HIFINet: Examination-Diagnosis-Treatment Hierarchical Feedback Interaction Network for Medication Recommendation

Medication combination recommendation is critical in clinic, since accurately predicting therapeutic drug can provide essential decision support to physicians. However, current approaches do not consider the multilevel structure of electronic health record (EHR) data or the hierarchical dependencies between multiple visits, leading to suboptimal recommendations. To address these limitations, we propose a novel hierarchical feedback interaction network (HIFINet) to utilize an examination-diagnosis-treatment hierarchical network for modeling the inherent multilevel structure of EHR data. The feedback long short-term memory network called FeLSTM, which is the basic unit of our hierarchical network, performs hierarchical interactions and leverages change information as feedback to propagate forward among different levels. Additionally, HIFINet contains four modules. First, an embedding module is designed to learn the health information representation of patients. Second, a three-layer time-series learning module is employed to capture temporal dependencies within each sequence. Next, a differential feedback interaction module is developed to capture the difference features between visits. Finally, an attention fusion module is used to learn a comprehensive representation of the patient’s health information and to recommend next multiple treatment medications. HIFINet is compared with state-of-the-art approaches on a real-world dataset. The results indicate that HIFINet outperforms other approaches, offering more accurate recommendations.

Repeated labile carbon inputs trigger soil microbial necromass decomposition through increasing microbial diversity and hierarchical interactions

Microbial necromass substantially contributes to soil organic carbon (SOC) sequestration. However, the response of soil microbial necromass to fresh labile carbon (C) inputs and the underlying microbial mechanisms are poorly understood. In this study, we investigated the dynamics of soil microbial necromass following single and repeated labile C inputs in two typical agricultural soils, black soil (Mollisol) and red soil (Ultisol). Our results showed that labile C inputs triggered the decomposition of soil native microbial necromass, regardless of soil type and C input frequency. Following repeated glucose pulses, the microbial necromass decreased by 75 % and 41 % in black and red soils, respectively, in a week. While a single glucose pulse triggered a comparable decline in soil microbial necromass, which occurred gradually over an incubation period of 3–4 weeks. In contrast, the priming effect following a single glucose pulse was 2–3 fold higher than repeated pulses. Repeated glucose pulses increased microbial α-diversity and temporal succession of various K-strategists, enhanced the network complexity and the potential hierarchical interactions between bacteria, fungi and protists, but yielded 1.66 fold less microbial biomass compared to a single pulse. These changes explained variations in soil microbial necromass after glucose pulses. Furthermore, repeated glucose pulses enhanced the linkages between microbial attributes and the degradation of soil microbial necromass. MicroResp assay revealed that the capacity of microbial communities to degrade microbially derived residues was 1.29 and 3.60 fold higher in black and red soils with repeated glucose pulses, respectively, compared to a single pulse. Our findings provide comprehensive insights into the microbially-mediated processes that influence soil microbial necromass degradation following labile C inputs, with important implications for understanding the dynamics and stabilisation of soil microbial necromass.

CEMINet: Context exploration and multi-level interaction network for salient object detection

Precise salient object detection (SOD) in realistic scenarios heavily relies on multi-scale context. Although existing SOD methods have made significant advancements by incorporating contextual information, they often overlook the correlation of contexts at different scales during feature extraction, leading to challenges in producing precise saliency maps. Our proposed solution to mitigate the above challenges associated with SOD is the Context Exploration and Multi-level Interaction Network (CEMINet). Specifically, we first develop a progressive multi-scale context extraction (PMCE) module, which enables the gradual capture of strongly correlated multi-scale context with the aid of multi-receptive-field convolution operations. Additionally, we design a hierarchical feature hybrid interaction (HFHI) module to effectively aggregate multi-level features by exploiting a hybrid interaction strategy following a top-down approach. And then, a well-designed stereoscopic attention enhancement (SAE) module is presented to refine multi-level features from HFHI through two parallel attention branches combined in a stereoscopic structure for generating precise predictions. Comprehensive experiments show our CEMINet achieves superior performance, without requiring any post-processing, compared to 16 state-of-the-art SOD models across five popular datasets. To substantiate the effectiveness and generality of our model, we also implement it to the camouflaged object detection (COD), outperforming the corresponding state-of-the-art models.

A human factor reliability analysis method for maritime transport based on an improved CREAM model and group decision-making

Maritime transportation has a crucial position in world trade, and maritime transportation accidents can cause large casualties, environmental pollution and economic losses. In order to evaluate human-caused and non-human-caused errors more comprehensively, so as to improve the reliability of maritime transportation, a reliability analysis model for maritime transporters is proposed, which combines quantitative and qualitative analysis, and can be used to analyze maritime transportation accidents retrospectively and predictively. Different from the traditional reliability analysis model, on the one hand, human-caused and non-human-caused events are analyzed by the optimized CREAM and GDM classifications respectively in the quantitative analysis part; on the other hand, the evolutionary process of the events and the logical hierarchical interactions of the sub-events can be analyzed in more detail by constructing the ETA-FTA analysis framework in the qualitative analysis part. In addition, in the optimization of the specific method, more attention is paid to the occurrence scenario and process of the event, so that the analysis results obtained through the model have a higher correlation with the actual data.

HeightFormer: A Multilevel Interaction and Image-Adaptive Classification–Regression Network for Monocular Height Estimation with Aerial Images

Height estimation has long been a pivotal topic within measurement and remote sensing disciplines, with monocular height estimation offering wide-ranging data sources and convenient deployment. This paper addresses the existing challenges in monocular height estimation methods, namely the difficulty in simultaneously achieving high-quality instance-level height and edge reconstruction, along with high computational complexity. This paper presents a comprehensive solution for monocular height estimation in remote sensing, termed HeightFormer, combining multilevel interactions and image-adaptive classification–regression. It features the Multilevel Interaction Backbone (MIB) and Image-adaptive Classification–regression Height Generator (ICG). MIB supplements the fixed sample grid in the CNN of the conventional backbone network with tokens of different interaction ranges. It is complemented by a pixel-, patch-, and feature map-level hierarchical interaction mechanism, designed to relay spatial geometry information across different scales and introducing a global receptive field to enhance the quality of instance-level height estimation. The ICG dynamically generates height partition for each image and reframes the traditional regression task, using a refinement from coarse to fine classification–regression that significantly mitigates the innate ill-posedness issue and drastically improves edge sharpness. Finally, the study conducts experimental validations on the Vaihingen and Potsdam datasets, with results demonstrating that our proposed method surpasses existing techniques.

Exploiting Hierarchical Interactions for Protein Surface Learning.

Predicting interactions between proteins is one of the most important yet challenging problems in structural bioinformatics. Intrinsically, potential function sites in protein surfaces are determined by both geometric and chemical features. However, existing works only consider handcrafted or individually learned chemical features from the atom type and extract geometric features independently. Here, we identify two key properties of effective protein surface learning: 1) relationship among atoms: atoms are linked with each other by covalent bonds to form biomolecules instead of appearing alone, leading to the significance of modeling the relationship among atoms in chemical feature learning. 2) hierarchical feature interaction: the neighboring residue effect validates the significance of hierarchical feature interaction among atoms and between surface points and atoms (or residues). In this paper, we present a principled framework based on deep learning techniques, namely Hierarchical Chemical and Geometric Feature Interaction Network (HCGNet), for protein surface analysis by bridging chemical and geometric features with hierarchical interactions. Extensive experiments demonstrate that our method outperforms the prior state-of-the-art method by 2.3% in site prediction task and 3.2 available at https://github.com/lyqun/HCGNet.

Hierarchical Graph Interaction Transformer With Dynamic Token Clustering for Camouflaged Object Detection.

Camouflaged object detection (COD) aims to identify the objects that seamlessly blend into the surrounding backgrounds. Due to the intrinsic similarity between the camouflaged objects and the background region, it is extremely challenging to precisely distinguish the camouflaged objects by existing approaches. In this paper, we propose a hierarchical graph interaction network termed HGINet for camouflaged object detection, which is capable of discovering imperceptible objects via effective graph interaction among the hierarchical tokenized features. Specifically, we first design a region-aware token focusing attention (RTFA) with dynamic token clustering to excavate the potentially distinguishable tokens in the local region. Afterwards, a hierarchical graph interaction transformer (HGIT) is proposed to construct bi-directional aligned communication between hierarchical features in the latent interaction space for visual semantics enhancement. Furthermore, we propose a decoder network with confidence aggregated feature fusion (CAFF) modules, which progressively fuses the hierarchical interacted features to refine the local detail in ambiguous regions. Extensive experiments conducted on the prevalent datasets, i.e. COD10K, CAMO, NC4K and CHAMELEON demonstrate the superior performance of HGINet compared to existing state-of-the-art methods. Our code is available at https://github.com/Garyson1204/HGINet.

A Multi-Dimensional Node Importance Evaluation Method Based on Graph Convolutional Networks

This paper addresses the problem of identifying, evaluating, and ranking key nodes in complex networks by introducing a novel Multi-Parameter Control Graph Convolutional Network (MPC-GCN) for assessing node importance. Drawing inspiration from the multidimensional and hierarchical interactions between nodes in physical systems, this approach integrates the automatic feature learning capabilities of Graph Convolutional Networks (GCNs) with a comprehensive analysis of nodes’ intrinsic properties, their interactions with neighbors, and their roles within the broader network. The MPC-GCN model offers an innovative framework for key node identification, leveraging GCNs to iteratively aggregate node and neighbor features across layers. This process captures and combines local, global, and positional characteristics, enabling a more nuanced, multidimensional assessment of node importance. Moreover, the model incorporates a flexible parameter adjustment mechanism, allowing the relative weights of different dimensions to be tuned, thereby adapting the evaluation process to various network structures. To validate the model’s effectiveness, we first tested the influence of model parameters on randomly generated small networks. We then conducted extensive simulations on eight large-scale networks using the Susceptible-Infected-Recovered (SIR) model. Evaluation metrics, including the M(R) score, Kendall’s tau correlation, the proportion of infected nodes, and the relative size of the largest connected component, were used to assess the model’s performance. The results demonstrate that MPC-GCN outperforms existing methods in terms of monotonicity, accuracy, applicability, and robustness, providing more precise differentiation of node importance. By addressing the limitations of current methods—such as their reliance on single-dimensional perspectives and lack of adaptability—MPC-GCN offers a more comprehensive and flexible approach to node importance evaluation. This method significantly improves the breadth and applicability of node ranking in complex networks.

Optimal policy for control of epidemics with constrained time intervals and region-based interactions

<p>We introduce a policy model coupled with the susceptible–infected- recovered (SIR) epidemic model to study interactions between policy-making and the dynamics of epidemics. We considered both single-region policies as well as game-theoretic models involving interactions among several regions and hierarchical interactions among policy-makers modeled as multi-layer games. We assumed that the policy functions are piece-wise constant with a minimum time interval for each policy stage, considering that policies cannot change frequently in time or be easily followed. The optimal policy was obtained by minimizing a cost function that consists of an implementation cost, an impact cost, and, in the case of multi-layer games, a non-compliance cost. We show, in a case study of COVID-19 in France, that when the cost function is reduced to the impact cost and parameterized as the final epidemic size, the solution approximates that of the optimal control in Bliman et al, (2021) for a sufficiently small minimum policy time interval. For a larger time interval, however, the optimal policy is a step down function, quite different from the step up structure typically deployed during the COVID-19 pandemic. In addition, we present a counterfactual study of how the pandemic would have evolved if herd immunity was reached during the second wave in the county of Los Angeles, California. Finally, we study a case of three interacting counties with and without a governing state.</p>

Nonuniform Defogging Based on Hierarchical Weight Interaction and Laplacian Prior

Nonuniform Defogging Based on Hierarchical Weight Interaction and Laplacian Prior

Hierarchical data synchronous interaction in nonlinear complex systems

Nonlinear complex systems are widely used in various industries, and the complexity and dynamism of the systems pose enormous challenges to data synchronization and interaction. This study proposes a method based on signal synchronization for data synchronization and interaction in nonlinear complex systems. The synchronization interaction path of data was analyzed, and signals were used as the carrier for data synchronization. The data synchronization interaction was carried out by combining the main synchronization signal and the auxiliary synchronization signal. Furthermore, mean filtering was introduced for signal filtering, and the main synchronization process was divided into two parts: coarse and fine tuning. The experimental results showed that the main synchronization frequency deviation of the research method was 303 Hz when the main synchronization completion rate was 100% in a −15 dB signal-to-noise ratio environment. In the analysis of data synchronization success rate, the research method achieved a highest success rate of 99.7% when the data transmission density was 50 pieces per minute. The experiment shows that the research method can effectively improve the quality and efficiency of data synchronization and interaction in nonlinear complex systems.

Polarimetric SAR Image Classification Based on Hierarchical Scattering-Spatial Interaction Transformer

Polarimetric SAR Image Classification Based on Hierarchical Scattering-Spatial Interaction Transformer

Tencent Text-Video Retrieval: Hierarchical Cross-Modal Interactions with Multi-Level Representations

Text-Video Retrieval plays an important role in multi-modal understanding and has attracted increasing attention in recent years. Most existing methods focus on constructing contrastive pairs between whole videos and complete caption sentences, while overlooking fine-grained cross-modal relationships, e.g., clip-phrase or frame-word. In this paper, we propose a novel method, named Hierarchical Cross-Modal Interaction (HCMI), to explore multi-level cross-modal relationships among video-sentence, clip-phrase, and frame-word for text-video retrieval. Considering intrinsic semantic frame relations, HCMI performs self-attention to explore frame-level correlations and adaptively cluster correlated frames into clip-level and video-level representations. In this way, HCMI constructs multi-level video representations for frame-clip-video granularities to capture fine-grained video content, and multi-level text representations at word-phrase-sentence granularities for the text modality. With multi-level representations for video and text, hierarchical contrastive learning is designed to explore fine-grained cross-modal relationships, i.e., frame-word, clip-phrase, and video-sentence, which enables HCMI to achieve a comprehensive semantic comparison between video and text modalities. Further boosted by adaptive label denoising and marginal sample enhancement, HCMI achieves new state-of-the-art results on various benchmarks, e.g., Rank@1 of 55.0%, 58.2%, 29.7%, 52.1%, and 57.3% on MSR-VTT, MSVD, LSMDC, DiDemo, and ActivityNet, respectively.

Tree species identity and interaction determine vertical forest structure in young planted forests measured by terrestrial laser scanning

Vertical forest structure is closely linked to multiple ecosystem characteristics, such as biodiversity, habitat, and productivity. Mixing tree species in planted forests has the potential to create diverse vertical forest structures due to the different physiological and morphological traits of the composing tree species. However, the relative importance of species richness, species identity and species interactions for the variation in vertical forest structure remains unclear, mainly because traditional forest inventories do not observe vertical stand structure in detail. Terrestrial laser scanning (TLS), however, allows to study vertical forest structure in an unprecedented way. Therefore, we used TLS single scan data from 126 plots across three experimental planted forests of a large-scale tree diversity experiment in Belgium to study the drivers of vertical forest structure. These plots were 9–11 years old young pure and mixed forests, characterized by four levels of tree species richness ranging from monocultures to four-species mixtures, across twenty composition levels. We generated vertical plant profiles from the TLS data and derived six stand structural variables. Linear mixed models were used to test the effect of species richness on structural variables. Employing a hierarchical diversity interaction modelling framework, we further assessed species identity effect and various species interaction effects on the six stand structural variables. Our results showed that species richness did not significantly influence most of the stand structure variables, except for canopy height and foliage height diversity. Species identity on the other hand exhibited a significant impact on vertical forest structure across all sites. Species interaction effects were observed to be site-dependent due to varying site conditions and species pools, and rapidly growing tree species tend to dominate these interactions. Overall, our results highlighted the importance of considering both species identity and interaction effects in choosing suitable species combinations for forest management practices aimed at enhancing vertical forest structure.

Transactive Energy System Deployment Over Insecure Communication Links

In this paper, the privacy and security issues associated with the transactive energy system (TES) deployment over insecure communication links are addressed. In particular, it is ensured that 1) individual agents’ bidding information is kept private throughout hierarchical market-based interactions; and 2) any extraneous data injection attack can be quickly and easily detected. An implementation framework is proposed to enable the cryptography-based enhancement of privacy and security for the deployment of any general hierarchical systems including TESs. Under the proposed framework, a unified cryptography-based approach is developed to achieve both privacy and security simultaneously. Specifically, privacy preservation is realized by an enhanced Paillier encryption scheme, where a block design is proposed to significantly improve computational efficiency. Attack detection is further achieved by an enhanced Paillier digital signature scheme, where a stamp-concatenation mechanism is proposed to enable detection of data replace and reorder attacks. Simulation results verify the effectiveness of the proposed cyber-resilient design for transactive energy systems. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This paper is motivated by addressing the issues of cyber resiliency for practically deploying transactive energy system (TES) but it is also applicable to the problem of enhancing the privacy and security for any general hierarchical control systems. TES is an emerging control approach that engages energy suppliers and customers through market operations and uses the price to optimally allocate energy resources. While it has been shown to be promising for power system applications, the underlying market-based interactions raise significant concerns of privacy (data leakage) and security (data tampering). However, existing TES works only focus on the coordination mechanism instead of privacy and security issues. This paper proposes a new cryptography-based TES design for practical deployment. Specifically, to protect privacy, individual supply and demand amounts to be exchanged are all encrypted in a particular way such that the original amounts cannot be inferred from the encrypted amounts, while the desired computation for setting the market clearing price can be carried out over the encrypted amounts, thus generating an encrypted result which, when decrypted, matches that of the same computation over the original amounts. To achieve security, for each exchanged data, its sender generates a particular digital signature which is exchanged together with the data. This enables the receiver to automatically detect the integrity by checking whether a mathematical relationship holds for the pair of data and signature. In our future research, we will investigate more challenging scenarios where some suppliers and customers themselves could be corrupted and purposely submit distorted amounts.

Joint role of corporate communication budget allocation and internet penetration in multinationals corporate brand equity cross-national effects: a hierarchical three-way interaction

Joint role of corporate communication budget allocation and internet penetration in multinationals corporate brand equity cross-national effects: a hierarchical three-way interaction

ПУБЛІЧНЕ УПРАВЛІННЯ ІННОВАЦІЙНИМ РОЗВИТКОМ РЕГІОНАЛЬНОГО АПК В КОНТЕКСТІ СИСТЕМНОГО ПІДХОДУ

The article examines the theoretical principles of public management of agro-industrial complex innovative development according to systemic approach. The purpose of the work is to develop the conceptual principles of using a systemic approach in the implementation of measures of public management of the innovative development of the agricultural sector. To achieve the goal, general scientific methods of knowledge and processes have been used: methods of induction and deduction - to clarify the interrelationship of the processes of public management of the agricultural sector innovative development; methods of system analysis and generalization - for systematization of existing scientific approaches, used theoretical sources and scientific literature; structuring method - to single out individual factors influencing the innovative development of the regional agricultural sector; method of system analysis, method of abstraction and formalization - for the development of conceptual foundations of the public administration system; graphic method - for displaying the theoretical and methodological material of the research. The prospects and possibilities of innovative development of agro-industrial complex enterprises have been determined. The importance of improving the mechanisms of public management in the direction of innovative development of agro-industrial complex has been substantiated. The main directions of public management of the innovative development of the agro-industrial complex have been considered. The conceptual provisions of the public management system of the agro-industrial complex innovative development have been formed. The generalized model of public management of the agricultural sector of the region based on the hierarchical interaction of key components has been proposed. The essence, purpose and strategic aspect of the development of the system of public management of the innovative development of the agro-industrial complex have been defined. The characteristics of the system of public management of the innovative development of the agro-industrial complex have been offered. It has been substantiated the expediency of introducing the concept of public-private partnership into the process of public management of the innovative development of the agro-industrial complex, which involves covering not only research organizations and producers, but also all participants in the production and sales chains of the agrarian sector.

DCMAI: A Dynamical Cross-Modal Alignment Interaction Framework for Document Key Information Extraction

Document key information extraction (DKIE) is a crucial topic that aims at automatically comprehending documents with complex formats and layouts (invoices, business insurance, etc.). While pre-trained approaches have shown high performance on many DKIE tasks, they suffer from three major challenges. First of all, these approaches ignore the ambiguity resulting from similar text representations before cross-modal interaction. Secondly, they do not consider cross-modal representation alignment before cross-modal interaction. Finally, self-attention layers in cross-modal interaction incur significant computing costs, making it hard to perform joint representation learning from all negative samples. To address these issues, we propose a Dynamical Cross-Modal Alignment Interaction framework (DCMAI). To be more specific, (1) A prior knowledge-guided module is designed to adaptively mine fine-grained visual information to disambiguate similar text representations. (2) A crossover alignment loss is formulated to align cross-modal representations before cross-modal interaction. (3) A hierarchical interaction sampling scheme is introduced to obtain a small but efficient subset of cross-modal negative samples, and a contrastive loss is employed to improve joint representation learning. Comprehensive experiments show that the proposed DCMAI achieves state-of-the-art performance than competitive baselines on several public downstream benchmarks. Code will be open to the public.

Особливості SMART-систем: необхідність і доцільність їх використання при моделюванні бізнес-процесів підприємства

Taking into account the key factors of the development of society, the expediency of using certain information technologies for the effective competitive development of enterprises is emphasized. The need for the use of appropriate information technologies is confirmed by the advantages of their implementation. In the course of the study, it is disclosed that the basis of the modern information society is certain SMART-systems, and the driver of competitive development of enterprise is the re-engineering of its business processes, which determined the purpose of the study. To achieve the goal of defining and clarifying the features of SMART-systems to ensure the effective implementation of business processes of the enterprise, the methods of analysis, synthesis, abstraction, comparison, idealization, analogy, and concretization were used. The content of the concept of «SMART-systems» is characterized, the features of the application are described. The expediency and necessity of dividing SMART-systems into certain types depending on their vector-level application have been determined. The features of SMART-systems of each of the selected groups are outlined. The expediency of dividing SMART-systems into the proposed groups is substantiated by the possibility of their more operative implementation, faster adaptation, wider range of application and the ability to combine them as constituent elements of the hierarchical structure of an integral system. The hierarchical interaction of information systems of different groups will contribute to the formation of a comprehensive SMART-system for ensuring the effective competitive development of an individual enterprise with its component technologies, software products, etc., due to the possibility of combining the advantages of some systems, leveling the difficulties of implementing others. It is substantiated that the integrity of the information and digital support of the enterprise’s activities will contribute to the choice of more effective options for the interaction of individual business processes, their more efficient implementation and more flexible editing depending on changes in the conditions for the functioning of enterprises, as well as focusing more attention on key business processes at a particular point in time.

A Multi-Task Transformer With Local-Global Feature Interaction and Multiple Tumoral Region Guidance for Breast Cancer Diagnosis.

Breast cancer, as a malignant tumor disease, has maintained high incidence and mortality rates over the years. Ultrasonography is one of the primary methods for diagnosing early-stage breast cancer. However, correctly interpreting breast ultrasound images requires massive time from physicians with specialized knowledge and extensive experience. Recently, deep learning-based method have made significant advancements in breast tumor segmentation and classification due to their powerful fitting capabilities. However, most existing methods focus on performing one of these tasks separately, and often failing to effectively leverage information from specific tumor-related areas that hold considerable diagnostic value. In this study, we propose a multi-task network with local-global feature interaction and multiple tumoral region guidance for breast ultrasound-based tumor segmentation and classification. Specifically, we construct a dual-stream encoder, paralleling CNN and Transformer, to facilitate hierarchical interaction and fusion of local and global features. This architecture enables each stream to capitalize on the strengths of the other while preserving its unique characteristics. Moreover, we design a multi-tumoral region guidance module to explicitly learn long-range non-local dependencies within intra-tumoral and peri-tumoral regions from spatial domain, thus providing interpretable cues beneficial for classification. Experimental results on two breast ultrasound datasets show that our network outperforms state-of-the-art methods in tumor segmentation and classification tasks. Compared with the second-best competitive method, our network improves the diagnosis accuracy from 73.64% to 80.21% on a large external validation dataset, which demonstrates its superior generalization capability.