Search

Network embedding, also known as graph embedding and network representation learning, is an effective method for representing graphs or network data in a low-dimensional space. Most existing methods focus on preserving network topology and minimizing the reconstruction errors to learn a low-dimensional embedding vector representation of the network. In addition, some researchers are devoted to the embedding learning of attribute networks. These researchers usually study the two matrices of network structure and network attributes separately, and then merge them to realize the embedding learning representation of attribute networks. These studies have different performances on a variety of downstream tasks. However, most of these methods have two problems: first, these methods mostly use shallow model to learn structure or attribute embedding, which do not make full use of the rich information contained in the network, such as the neighborhood information of nodes; second, the distribution of the learned network low-dimensional vector representation is overlooked, which leads to poor generalization ability of the model in some real-world network data. Therefore, this paper proposes an adversarially regularized network representation learning model based on attention mechanism, which encodes the topology features and content information of the network into a low-dimensional embedding vector representation through a graph attention autoencoder. Meanwhile, through an adversarial training schema, the learned low-dimensional vector representation could circumvent the requirement of an explicit prior distribution, and thus obtain better generalization ability. Extensive experiments on tasks of link prediction and node clustering demonstrate the effectiveness of learned network embeddings.

Micro/nanomotors (MNMs) have attracted tremendous interest of researchers owing to their ability to mimic the function of natural microorganisms to complete tasks in biomedicine and other fields. However, in order to realize their practical use in the biomedical field, MNMs are required to meet the biosafety standard, which has not been systematically reviewed. The most important thing of concern for biosafety is to fabricate MNMs with good biodegradability, biocompatibility and biosafe propulsion. We reviewed the state-of-the-art of biosafe MNMs, focusing on the biodegradability, biocompatibility, biosafe propulsion and biosafety assessments of MNMs. The challenges and opportunities of biosafe MNMs for biomedical application are discussed.

Considering potential applications in power systems, economic dispatch problem (EDP) in time-varying balanced communication networks is studied, aiming to minimise the total cost of generating electricity. This is equivalent to dealing with the problem of optimising the sum of local functions while a single generator possesses only its local function. The variables of generators, satisfying some local constraints, are coupled by a linear constraint. In order to resolve the EDP, the authors design a fully distributed primal-dual optimisation algorithm with time-varying uncoordinated step-sizes. In consideration of saving computation and communication resources, an event-triggered scheme is introduced into the algorithm, based on which each generator is only allowed to interact with their neighbouring generators at some independent event-triggered sampling time instants. The proposed algorithm is able to achieve a linear convergence rate under the strong convexity and smoothness of the local objective functions. The Zeno-like behaviour is rigorously excluded, which means that the interval between any two consecutive sampling time instants of each generator is not less than two. Effectiveness of the algorithm and correctness of the theoretical analysis are verified by numerical experiments.

With the development of hydropower stations towards multi-objective development, such as water supply, flood control, power generation and navigation, rational allocation of water resources has become a major problem. Based on liuzhou Honghua Hydropower station, an optimal scheduling model with multi-constraint, multi-objective, and multi-variable hydropower stations is proposed supported by the objective function based on the maximum generating capacity, navigation capacity and the least breakdown power. The Pareto optimal solution is employed to solve the multi-objective problem of optimal scheduling of hydropower stations. Consequently, a comprehensive evaluation model for optimal scheduling of hydropower stations was established according to the TOPISIS and gray correlation degree analysis. By selecting the scheduling solutions of hydropower stations, the result could be more accurate and closer to real conditions, which could provide guidance for project decision-makers.

Relying on the remote sensing image of GF-1,GF-2, DEM datas, GIS datas and other datas resources of nature reserve, this paper develops a dynamic monitoring system for nature reserves of B/S version, based on Web RS, Web GIS, machine learning, statistical analysis and big datas technology organically integrated. After a series of preprocessing and interpretation of multi-temporal remote sensing image datas of nature reserve, the system can show changes of land use, vegetation cover and human activity monitoring elements in nature reserve. It is concluded that the accuracy, reliability and robustness of remote sensing image classification are improved by making full use of different disciplines theories and technologies, so that the land classification information of nature reserves can be identified and extracted more quickly and accurately, which can be used for land use planning, land use structure adjustment and government decision-making of nature reserves, and provides a forward method for effectively solving the main scientific problems and key technical problems of land use management in nature reserves under changing environment.

The paper studies finite-time quasi-synchronization (FTQS) for time-delay multilayer networks (MLNs), the FTQS of MLNs is achieved by applying delay controller. Compared with some existing results, the optimal combination relevance between exponential power of FT controller and initial value of system is gained when the QS time is minimal. In particular, there is no necessary correlation between the FT convergence of the system and the range of QS error. Lastly, the validity of the advanced method is verified by FTQS simulation of MLNs with financial oscillators.

Pb(II), a highly toxic heavy metals that targets the human nervous system, has contaminated water and food in recent years. The nanomaterial of Metal–organic frameworks (MOFs) possesses superior removal performances. However, their practical application is limited by their nature of fine powdery. Herein, we developed robust MOFs film without the adding of polymer in MOFs ink and the modification of the substrate, which was achieved by choosing appropriate substrate of gelatin hydrogel with functional groups and simple thermal-treatment after coating UiO-66-NO2 onto gelatin hydrogel. The volatilization of the UiO-66-NO2 dispersant ethyl acetate promotes the rearrangement and dense accumulation of UiO-66-NO2 crystals to form a dense UiO-66-NO2 film. During the following thermal-treatment process, –COOH of UiO-66-NO2 and –NH2 of gelatin chemically reacted to establish the stable composites of UiO-66-NO2 film and gelatin hydrogel. The results of batch adsorption experiments demonstrate that adsorption behaviors of the as-synthesized gelatin/UiO-66-NO2 film composite hydrogels can be explained by the pseudo-second-order kinetic model and the Langmuir isotherm model, stating that this adsorption process involves with a chemical monolayer adsorption. Besides, gelatin/UiO-66-NO2 film composite hydrogels own satisfactory removal efficiency for Pb(II) in apple juice, and the adsorption process has little impairment on the quality of apple juice. These results suggest that MOFs film is suitable for water decontamination and food safety control fields.

A physical-layer network coding (PNC) based inter-ONU-communication (IOC) scheme is proposed for next generation high-speed PONs which apply four-level pulse amplitude modulation (PAM4). A 25 Gb/s full-duplex and private communication is demonstrated.

The E3 ubiquitin protein UBR5 has been implicated in the regulation of multiple biological functions and has recently emerged as a key regulator of the ubiquitin-proteasome system (UPS) in cancer. However, the clinical significance and biological function of UBR5 in colorectal cancer (CRC) are poorly understood. In this study, we compared the expression pattern of UBR5 between CRC and adjacent normal tissues and found that UBR5 expression was frequently elevated in CRC, possibly through chromosomal gains. Using three CRC patient cohorts, we found that patients with high UBR5 mRNA levels, UBR5 gene amplification, or high nuclear UBR5 protein levels had poor prognoses. Multivariate analysis showed that the alterations in UBR5 were independent predictors of CRC prognosis with the TNM stage as a confounding factor. Furthermore, knockdown of UBR5 prevented the proliferation, colony formation, migration, and invasion of CRC cells in cell culture models. An in vivo animal model further confirmed that UBR5 knockdown reduced the growth of CRC tumors. In conclusion, our study is the first to systematically investigate the clinical and biological significance of UBR5 and to conclude that an elevated UBR5 level plays an oncogenic role and may be a potential prognostic marker in CRC.

To explore the complex molecular mechanisms of bladder cancer, mRNA and miRNA expression profiles were combined for systematic analyses. A total of 18 common differentially expressed genes (DEGs) were identified from two mRNA expression datasets which consisted of 206 tumor and 74 normal tissues. Then, survival analysis based on the SurvExpress database showed that the common DEGs were able to significantly differentiate low- and high-risk groups in 4 public bladder cancer datasets (p

This paper investigates the distributed optimization problem over multi-agent networks, in which the target of agents is to collaboratively optimize the sum of all local objective functions. Each local objective function is uniquely known by a single agent. We concentrate on the scenario where communication among agents is portrayed as directed graphs. Based on the exact first order method, a fully distributed optimization algorithm is proposed to deal with the optimization problem. The proposed algorithm utilizes row-stochastic matrices and uncoordinated step-sizes, which exactly drives all agents to converge to the global optimization solution. Under the assumptions that the global objective function is strong convex and the local objective functions have Lipschitz continuous gradient, we show that the proposed algorithm linearly converges to the global optimization solution as long as the maximum step-size of agents does not exceed an explicitly characterized upper bound. Finally, numerical experiments are presented to demonstrate the correctness of theoretical analysis.

The prediction and calculation of water resource demand play a crucial role in the balance of water resource supply and demand in the research area. Only when the water demand of the research area is accurately and reasonably grasped, it is possible to make corresponding management and adjustment according to the actual situation in the process of water resource utilization, so as to ensure the sustainable development and utilization of water resource. In this paper, the upper and middle reaches of the Pearl River basin is taken as the research area. According to the relationship between supply and demand and the social and economic situation of the Pearl River basin, VENSIM PLE is used to establish the dynamic model of the water demand forecasting system of the Pearl River Basin. The model involves four subsystems of population, industry, agriculture and water resource, and the water resource demand of the three provinces is simulated. Compared with the results of the pearl River Basin planning report, the prediction error is small, which proves that the system dynamics model is feasible to predict the water demand of the Pearl River Basin.

Tumor boundary delineation using F-FDG PET/CT is a promising tool for radiotherapy applications, but no consensus has been established regarding the optimal delineation method. Time-phase variability of F-FDG PET/CT imaging frequently affects metabolically active volumes and treatment planning for nasopharyngeal carcinoma (NPC). This study aimed to evaluate the time-phase robustness of 8 methods commonly used for tumor volume delineation in NPC.Twenty patients with biopsy-proven NPC were included and underwent multiple time-phase F-FDG PET/CT imaging. Gross tumor volumes (GTVs), absolute SUV, gradient-based watershed segmentation (GTV-GWT), and anatomic biologic contouring (GTV-ABC) values were determined. The volume of overlap between GTV-CT and the 8 PET-based GTVs was enclosed and the overlap fraction (OF-CT) calculated. Color matrix was used to semiquantify the time-phase differences. Gross tumor volume values obtained with different methods were recorded and compared using paired t test. Time-phase differences of GTVs and SUVmax were compared among groups by analysis of variance with Tukey honest significance tests. The coefficients of variation were computed to assess intrapatient time-phase variability. Similarity coefficient was calculated to evaluate similarity.Differences were observed between GTVs obtained at different time points using various delineation procedures. Nonsignificantly higher percentages were obtained for GTV-GWT (88.17%) and GTV-ABC (86.98%) compared with other methods, showing their robustness. GTV-40% (0.81-0.88) and GTV-ABC (0.82-0.88) indicated higher similarity with GTV-MRI than the other methods.PET/CT-based GTV-ABC between 35 and 55 minutes should be the first choice for NPC treatment planning.

Many foreign firms tend to follow the market-seeking mandate in China. However this orientation alone does not guarantee superior performance. From the perspectives of strategic fit and institutional theory, this research seeks to reveal several conditions under which market-seeking MNEs can achieve superior performance in China. We identify three performance contributors to marketing seeking FDI: the host country’s favorable formal institutions towards FDI, the subsidiaries’ operational experience and absorptive capacity in the host country, and the ownership structure of the subsidiary. Using data of 5,080 foreign invested subsidiaries in 2003–2010, our findings support the hypotheses that market-seeking orientation becomes more profitable for foreign subsidiaries in China when (1) the host country provides a more favorable institutional framework towards FDI; (2) the subsidiary has a longer history of FDI operation in the host country that leads to stronger absorptive capacity; and (3) the subsidiary is organized in a wholly owned manner.

Non-alcoholic fatty liver disease is highly associated with morbidity and mortality in population. Although studies have already demonstrated that the immune response plays a pivotal role in the development of non-alcoholic fatty liver disease, the comprehensive regulation is unclear. Therefore, present study was carried out to investigate the non-alcoholic fatty liver disease development under neutrophil depletion. To achieve the aim of the study, C57BL/6 J mice were fed with high fat diet for 6 weeks before treated with neutrophil deplete antibody 1A8 or isotype control (200 μg/ mouse every week) for another 4 weeks. Treated with 1A8 antibody, obese mice exhibited better whole body metabolic parameters, including reduction of body weight gain and fasting blood glucose levels. Neutrophil depletion also effectively reduced hepatic structural disorders, dysfunction and lipid accumulation. Lipid β-oxidative markers, phosphorylated-AMP-activated protein kinase α and phosphorylated-acetyl-CoA carboxylase levels were increased in 1A8 antibody-treated obese mouse group. The mitochondrial number and function were also reversed after 1A8 antibody treatment, including increased mitochondrial number, reduced lipid oxidative damage and enhanced mitochondrial activity. Furthermore, the expression of inflammatory cytokines, tumor necrosis factor-α, interleukin-6, and monocyte chemoattractant protein-1 were obviously reduced after neutrophil depletion, accompanied with decreased F4/80 mRNA level and macrophage percentage in liver. The decreased NF-κB signaling activity was also involved in the beneficial effect of neutrophil depletion. Taken together, neutrophil depletion could attenuate metabolic syndromes and hepatic dysfunction.