Search

With the rapid development of Internet social media,using recommendation algorithms to effectively model and filter massive amounts of information has become the key to predict user behavior preferences,hot spot tendency,network security si-tuation and other issues.At the same time,with the development of deep learning,graph neural network model has achieved good results in solving the dense graph structure data in recommendation system.Collaborative filtering algorithm,as the most widely used recommendation algorithm,uses user-item group interaction data to predict users' future preferences and item ratings.However,existing recommendation algorithms still face the problems of data sparseness and cold start,and lack of a good quantification of uncertainty.This paper proposes an inductive matrix completion prediction fusion algorithm based on probabilistic meta-learning(MetaIMC),which re-characterizes meta-learning from the perspective of Bayesian inference,builds a robust GNN-meta-learning model,and makes full use of data priors to build solutions for learning new tasks from sparse data.Firstly,MetaIMC can effectively use variational Bayesian inference to obtain the prior distribution,alleviate the uncertainty and ambiguity in the meta-model task training,and further improve the generalization ability of the model.Secondly,MetaIMC can implement new user reco-mmendations and solve the cold start problem without any user side information.Finally,in the two scenarios of traditional matrix completion and user cold start,the performance of the model is evaluated by using three public datasets of Flixster,Douban and Yahoo_music,which verifies the effectiveness of MetaIMC on traditional matrix completion task,and achieves the best performance on the cold start problem.

Tailoring the specific properties to practical applications by structural modification is of vital importance for the envisioned development of two-dimensional ferroelectric materials. Herein, a comprehensive investigation on the effects of single doping on the ferroelectric properties and electronic transport in a monolayer of α-In2Se3 was carried out via the combination of first-principles density functional theory calculations and electron–phonon coupling simulations. Our results show that single-doping in In2Se3 can reduce effective mass of carriers and thereby enhance the high carrier mobility potential of the material. Moreover, the ferroelectric phonon mode in single-doped In2Se2X features a lower scattering rate, associating with the single-doping atom, and indicates reduced hindrance to carrier transport during ferroelectric switching. Compared to pristine In2Se3, the obtained smaller ferroelectric barriers (<1 eV) of single-doped ones promote more promising ferroelectricity from the analysis of the ferroelectric soft mode. Interestingly, the observed variations in ferroelectric behaviors resulting from doping of different elements highlight the significance of single-doping in modifying the ferroelectric properties of monolayers. Furthermore, strain engineering results reveal that single doping obviously affects the dependence of gap on strains: linear relationship for doping ones and nonlinearity for pristine one. Our study provides valuable insights into achieving higher carrier mobility in these critical materials. [ABSTRACT FROM AUTHOR]

Objectives: The application of rifaximin, a non‐absorbable antibiotic, in hepatic encephalopathy (HE) has been well established; however, its effect on other complications in cirrhotic patients with previous gastroesophageal variceal bleeding (GEVB) remains unclear. Therefore, we performed a pilot randomized controlled trial aiming to evaluate the impact of rifaximin on cirrhosis‐related complications and changes in gastric microbiota. Methods: Eighty cirrhotic patients who received prophylactic endoscopic treatment for variceal rebleeding were randomly assigned to the control or rifaximin treatment group (rifaximin 400 mg twice daily for 8 weeks). Primary outcome was the total liver‐related score, consisting of changes in cirrhosis‐related complications including rebleeding, ascites, HE and portal vein thrombosis (PVT). The 16S rDNA sequencing analysis was conducted with gastric lavage fluid samples for the analysis of gastric microbiota. Results: During the 8‐week follow‐up, the total liver‐related score decreased significantly upon rifaximin therapy (−0.35 ± 0.14 vs 0.05 ± 0.14, p = 0.0465) as well as serum C‐reactive protein (CRP) (p = 0.019) and interleukin‐8 (p = 0.025) compared with the control group. The rate of PVT recanalization was significantly higher in the rifaximin group (p = 0.012). Prominent difference in gastric microbiota between the two groups was observed, and the rifaximin group had a higher abundance of several taxa which were dysregulated in the progression of cirrhosis. CRP was correlated with several taxa including Alphaproteobacteria, Rhizobiales and Collinsella. Conclusions: Rifaximin may improve cirrhosis‐related complications, including PVT, in patients with previous GEVB through anti‐inflammatory and microbiota‐modulating functions. Trial registration number: NCT02991612. [ABSTRACT FROM AUTHOR]

In this study, an electrode slurry composed of molybdenum disulfide (MoS2) and vapor-grown carbon fiber (VGCF) prepared through a solid-phase synthesis method was blade-coated onto copper foil to form a thick film as the anode for lithium-ion batteries. In previously reported work, MoS2-based lithium-ion batteries have experienced gradual deformation, fracture, and pulverization of electrode materials during the charge and discharge cycling process. This leads to an unstable electrode structure and rapid decline in battery capacity. Furthermore, MoS2 nanosheets tend to aggregate over charge and discharge cycles, which diminishes the surface activity of the material and results in poor electrochemical performance. In this study, we altered the density of the MoS2–carbon fiber/Cu foil anode electrode by rolling. Three different densities of electrode sheets were obtained through varying rolling repetitions. Our study shows the best electrochemical performance was achieved at a material density of 2.2 g/cm3, maintaining a capacity of 427 mAh/g even after 80 cycles. [ABSTRACT FROM AUTHOR]

Aim: Genome‐wide association studies (GWAS) have identified multiple susceptibility loci associated with insulin resistance (IR)‐relevant phenotypes. However, the genes responsible for these associations remain largely unknown. We aim to identify susceptibility genes for IR‐relevant phenotypes via a transcriptome‐wide association study. Materials and Methods: We conducted a large‐scale multi‐tissue transcriptome‐wide association study for IR (Insulin Sensitivity Index, homeostasis model assessment‐IR, fasting insulin) and lipid‐relevant traits (high‐density lipoprotein cholesterol, triglycerides, low‐density lipoprotein cholesterol and total cholesterol) using the largest GWAS summary statistics and precomputed gene expression weights of 49 human tissues. Conditional and joint analyses were implemented to identify significantly independent genes. Furthermore, we estimated the causal effects of independent genes by Mendelian randomization causal inference analysis. Results: We identified 1190 susceptibility genes causally associated with IR‐relevant phenotypes, including 58 genes that were not implicated in the original GWAS. Among them, 11 genes were further supported in differential expression analyses or a gene knockout mice database, such as KRIT1 showed both significantly differential expression and IR‐related phenotypic effects in knockout mice. Meanwhile, seven proteins encoded by susceptibility genes were targeted by clinically approved drugs, and three of these genes (H6PD, CACNB2 and DRD2) have been served as drug targets for IR‐related diseases/traits. Moreover, drug repurposing analysis identified four compounds with profiles opposing the expression of genes associated with IR risk. Conclusions: Our study provided new insights into IR aetiology and avenues for therapeutic development. [ABSTRACT FROM AUTHOR]