Bo Chen

Causal discovery of type 2 diabetes mellitus (T2DM) risk factors is crucial for revealing underlying disease mechanisms and supporting prevention and treat

Most existing cascading failure models for command and control networks (C2Ns) primarily focus on node-level failures while overlooking edge-level failures. To address this issue, this paper proposes a novel edge-based cascading failure model that integrates edge importance and hierarchical structure. First, both local and global edge importance are quantified through edge degree and ego network edge betweenness, and an improved definition of edge importance is introduced by incorporating an enhanced bridging coefficient. Second, based on the hierarchical structure of network nodes, edge hierarchy is defined, and a new initial edge load formulation is proposed by combining edge hierarchy and importance. Third, to reflect load imbalance in realistic networks, a nonlinear edge capacity model is constructed, and a non-uniform, adjustable load redistribution strategy based on the residual capacity of neighboring edges is developed to accommodate network hierarchy. Simulation experiments are conducted to validate the proposed model. The results show that, in two representative C2N networks, the degradation rates of average efficiency and connectivity coefficient are 22.42% and 73.70% for Network 1, and 13.84% and 14.52% for Network 2, respectively. These results demonstrate that the proposed model significantly improves the robustness and resilience of C2Ns under cascading failure scenarios.

This paper presents an improved unmanned aerial vehicle (UAV) path planning algorithm that combines the advantages of fuzzy control and the artificial potential field (APF) algorithm. This hybrid approach addresses the limitations of the rapidly‐exploring random tree star (RRT*) algorithm, which suffers from poor search path quality and low search efficiency. We introduce a goal‐biased strategy that divides the sampling process of RRT* into two stages: exploration and convergence. A dual‐input single‐output fuzzy controller dynamically adjusts the goal‐bias based on obstacle changes and the distance to the target. The paper also proposes a directed node generation approach to enhance the algorithm's adaptability and purposefulness during the exploration stage. The APF algorithm is integrated to guide the random generation of the new node, improving the overall efficiency. The results of simulation experiments conducted in two environments show that the proposed algorithm outperforms other algorithms in terms of higher path quality, fewer redundant nodes, and faster search speeds.

In the context of information warfare, command and control (C2) networks are exhibiting increasingly prominent multi-network dependency characteristics, leading to a growing interest in the study of interdependent C2 network models. We propose a modeling approach for interdependent C2 networks based on integrated reconnaissance-strike and OODA loop, addressing the limitations of existing edge connection strategies in effectively capturing the interdependent coupling relationships within the network. First, the interdependent relationships within the network are described, and a dual-layer structural model of interdependent C2 networks is abstracted. Second, based on the local efficiency of nodes, an edge connection strategy for sensing and firepower nodes under integrated reconnaissance-strike is proposed. Third, interdependence strength and link balance are defined, and an inter-layer coupling edge connection strategy based on interdependence strength and link balance is proposed. Finally, algorithmic simulations are designed to analyze the network properties of the model and network performance under different edge connection strategies. Simulation results demonstrate that the proposed modeling method effectively captures the interdependent characteristics of C2 networks while exhibiting enhanced network resilience against destruction.

Network public opinion is the reflection of social public opinion on the internet and an important part of social public opinion. There are various forms of network public opinion, including WeChat, Weibo, Twitter, etc., and the comments and interactions on these multiple platforms constitute the rich content of network public opinion. Real-time understanding of network public opinion propagation helps to grasp public feedback and emotions. When a public opinion crisis occurs, adopting effective control methods can quickly respond to and handle social hot events, avoid misunderstandings and the expansion of contradictions, and help maintain social harmony and stability. To address this problem, we propose a public opinion propagation model based on DI-SCIR for two-layer coupled social network. A strong coupled inter-layer connection method based on LBRank is proposed to establish a practical two-layer coupled social network. Theoretical analysis is conducted on the impact of different factors in the model on public opinion dissemination of coupled social network. A control method of direct immune SCIR is designed, which intervenes with real-time online users during the spreading of public opinion. This can guide public opinion to spread in a favorable direction. The simulation results show that the constructed public opinion propagation model for two-layer social network can better reflect the spreading of public opinion in real life. The public opinion control method can quickly guide and intervene in the spread of public opinion and reduce the impact of negative public opinion.

Due to the inadequacy of edge importance identification methods for single network connectivity or transmission capability, a novel approach to edge importance ranking: balancing network structure and transmission performance is proposed. First of all, a new definition for network energy intensity and network constraint coefficient is presented, and an improved structure hole method is proposed and verified. It solves the problem that the traditional structural hole method cannot be used to identify important edges directly. Second, the edge information entropy of the command and control network is defined by introducing the ego-network betweenness centrality, and the complexity of information entropy calculation is reduced. Furthermore, the novel edge importance ranking method based on the combination of the improved structure hole and information entropy is proposed. This method solves the problem of edge importance evaluation which takes into account network structure and transmission performance. Finally, the performance simulation experiments are designed based on different network models and edge importance ranking methods. Simulation results show that the proposed method has higher recognition accuracy and better universality for important edges.

KSII Transactions on Internet and Information Systems TIIS Paper Details

Statistical correlation analysis is currently the most typically used approach for investigating the risk factors of type 2 diabetes mellitus (T2DM). However, this approach does not readily reveal the causal relationships between risk factors and rarely describes the causal relationships visually. Considering the superiority of reinforcement learning in prediction, a causal discovery approach with reinforcement learning for T2DM risk factors is proposed herein. First, a reinforcement learning model is constructed for T2DM risk factors. Second, the process involved in the causal discovery method for T2DM risk factors is detailed. Finally, several experiments are designed based on diabetes datasets and used to verify the proposed approach. The experimental results show that the proposed approach improves the accuracy of causality mining between T2DM risk factors and provides new evidence to researchers engaged in T2DM prevention and treatment research.

Considering the shortcomings of the existing network key node identification methods based on multiattribute fusion, which have single evaluation methods and low decision accuracy, combined with the advantages of the high accuracy of TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) algorithm and the applicability of grey relational analysis method for incomplete information evaluation, the concept of relative closeness is proposed, and nodes are ranked in importance based on the relative closeness; a key node identification method algorithm based on improved multiattribute fusion is designed. First, the identification problem of key nodes is transformed into multiattribute decision-making method, and the decision matrix is obtained. Second, the weighting matrix is obtained by weighting them in both subjective and objective dimensions, the relative closeness is calculated for the weighting matrix. Finally, sort the network nodes by relative closeness, and network performance simulation experiments are designed using various combinations of evaluation methods and key node identification methods. The simulation results show that this method is more adaptable and improves the identification accuracy of the network key nodes.