Search

BackgroundPatients with COVID-19 in the intensive care unit (ICU) have a high mortality rate, and methods to assess patients’ prognosis early and administer precise treatment are of great significance. ObjectiveThe aim of this study was to use machine learning to construct a model for the analysis of risk factors and prediction of mortality among ICU patients with COVID-19. MethodsIn this study, 123 patients with COVID-19 in the ICU of Vulcan Hill Hospital were retrospectively selected from the database, and the data were randomly divided into a training data set (n=98) and test data set (n=25) with a 4:1 ratio. Significance tests, correlation analysis, and factor analysis were used to screen 100 potential risk factors individually. Conventional logistic regression methods and four machine learning algorithms were used to construct the risk prediction model for the prognosis of patients with COVID-19 in the ICU. The performance of these machine learning models was measured by the area under the receiver operating characteristic curve (AUC). Interpretation and evaluation of the risk prediction model were performed using calibration curves, SHapley Additive exPlanations (SHAP), Local Interpretable Model-Agnostic Explanations (LIME), etc, to ensure its stability and reliability. The outcome was based on the ICU deaths recorded from the database. ResultsLayer-by-layer screening of 100 potential risk factors finally revealed 8 important risk factors that were included in the risk prediction model: lymphocyte percentage, prothrombin time, lactate dehydrogenase, total bilirubin, eosinophil percentage, creatinine, neutrophil percentage, and albumin level. Finally, an eXtreme Gradient Boosting (XGBoost) model established with the 8 important risk factors showed the best recognition ability in the training set of 5-fold cross validation (AUC=0.86) and the verification queue (AUC=0.92). The calibration curve showed that the risk predicted by the model was in good agreement with the actual risk. In addition, using the SHAP and LIME algorithms, feature interpretation and sample prediction interpretation algorithms of the XGBoost black box model were implemented. Additionally, the model was translated into a web-based risk calculator that is freely available for public usage. ConclusionsThe 8-factor XGBoost model predicts risk of death in ICU patients with COVID-19 well; it initially demonstrates stability and can be used effectively to predict COVID-19 prognosis in ICU patients.

Building emissions reduction has become a crucial pathway for China to achieve its ‘dual-carbon’ goals. As an integrated energy entity coupled with multi-energy flow networks， smart buildings face challenges such as high carbon emissions， a high degree of coupling in multi-energy flow networks， and distinct dynamic characteristics in load energy consumption behavior. In response to these challenges， a low-carbon scheduling method for multi-energy flow buildings based on deep reinforcement learning （deep RL） is proposed. Firstly， a reward and punishment ladder-type carbon emissions trading mechanism is established based on the actual carbon emissions of smart buildings. Secondly， targeting the carbon market and multi-energy flow coupling networks， a low-carbon scheduling model for multi-energy flow buildings is developed， aiming to minimize operating costs as the objective function， and the scheduling is transformed into a Markov decision process （MDP）. Subsequently， the Rainbow algorithm is employed to solve the optimal scheduling. Finally， the feasibility and effectiveness of the optimal scheduling model are verified through simulation analysis.

As an integrated energy network with multi-energy flow interconnected， virtual power plants （VPPs） face difficulties such as high actual carbon emissions， insufficient new energy consumption capacity， complex optimization variables of multi-energy flow， and non-uniform optimization objectives. To this end， a multi-objective optimal scheduling method for virtual power plants considering carbon capture and multi-energy flow is proposed to improve the new energy consumption capacity of multi-energy flow virtual power plants and reduce carbon emissions. First， a reward and punishment ladder-type carbon trading mechanism is established based on the actual carbon emissions of the virtual power plant. Then， a dual-objective optimal scheduling model is built with the minimization of operating costs and carbon emissions as the objective function and solved by using the multi-objective salp swarm algorithm. Finally， multiple scheduling scenarios are set up to compare and analyze the performance of the multi-objective optimization algorithm. The results verify the feasibility and effectiveness of the optimal scheduling model.

Lactic acid bacteria and yeast are abundant in kefir grains. Six natural kefir grains were used in this research to isolate and purify 64 strains of yeast and 108 strains of lactic acid bacteria. A total of three lactic acid bacteria and one yeast strain (Lactococcus lactis 2C6, Lactobacillus helveticus 6171, Lactobacillus plantarum 4M2 and Saccharomyces cerevisiae 6Y6) were examined for their ability to protein hydrolysis, water holding capacity, alcohol production, acid and bile salt tolerance and identified by 16S rDNA. The proteolytic ability, alcohol generation, and acid and bile salt tolerance of 4M2 and 6Y6 were excellent. Lyophilization was used to create bacterial powder, which laid the foundation for the subsequent development of direct-vat-set (DVS) kefir starter.

[Purpose/Significance] Intellectual capital is the important intangible asset and core competitiveness of university affiliated think tanks, and studying its intellectual capital acquisition mechanism is of great significance to build high-level university affiliated think tanks. Through literature review, this paper makes clear the components and access of intellectual capital of university affiliated think tanks. [Method/Process] Based on the theory of innovation-driven development, this paper constructs the model of innovation-driven intellectual capital acquisition mechanism of university affiliated think tanks, and applies this model to analyze the current situation of intellectual capital acquisition mechanism of 100 famous university affiliated think tanks at home and abroad. [Results/Conclusions] It shows that the model of innovation-driven intellectual capital acquisition mechanism of university affiliated think tanks consists of human capital, structural capital and relational capital. Human capital acquisition is the main way for university affiliated think tanks to acquire intellectual capital, and structural capital acquisition and relational capital acquisition, as auxiliary ways, play roles in safeguarding and promoting to acquire human capital. Moreover, the countermeasures and suggestions for building the first-class university affiliated think tanks in China are: to form an effective model of human capital acquisition, to establish an improved access channel of structural capital acquisition, and to build a three-dimensional system of relational capital acquisition.

Mechanisms underlying large‐scale spatial patterns of species richness are one of the central issues in ecology. Although contemporary climate, evolutionary history, and historical climate change have been proposed as drivers of species richness patterns, variation in the relative importance of different factors remains a major challenge. Here, using newly compiled distribution data with a spatial resolution of 100 × 100 km for 43 023 angiosperms plant species in east Eurasia, we mapped species richness patterns for plants with different growth forms (i.e. woody versus herbaceous) and range sizes (i.e. wide‐ranged versus narrow‐ranged species), and compared the relative importance of the four hypotheses in explaining these patterns, i.e. freezing tolerance hypothesis, historical climate change hypothesis, Janzen's hypothesis (predicting that climate seasonality and topography determine species richness patterns), and diversification rate hypothesis. We found that species richness of all angiosperm plants presented a clear latitudinal gradient and was highest in southwestern China and Central Asian mountains. Notably, species richness patterns and their dominant drivers differed between species groups. Historical climate change was the dominant driver for richness patterns of all and herbaceous species. The freezing tolerance hypothesis dominated the drivers for all woody species, while Janzen's hypothesis dominated narrow‐ranged woody and herbaceous species. Our study suggests that different hypotheses contribute to large‐scale plant richness patterns via their effects on different plant groups. Our results did not support the diversification rate hypothesis, but demonstrated the high importance of historical climate change to plant diversity in east Eurasia, providing new perspectives on the mechanisms of plant diversity patterns in this continent. [ABSTRACT FROM AUTHOR]

ABSTRACT Aim Location Major Taxa Studied Methods Results Main Conclusions The Kunming‐Montreal Global Biodiversity Framework requires that evolutionary histories of species should be considered in conservation planning. The phylogenetic structure of species assemblages quantifies species evolutionary histories and increasingly becomes an endeavour for ecologists. Understanding the geographic patterns of phylogenetic structure of species assemblages and their drivers can provide a fundamental reference for conservation planning. Although several theoretical hypotheses based on the effects of contemporary environment, historical climate change and evolutionary niche conservatism had been widely discussed in previous studies, the relative contributions of these hypotheses on phylogenetic structure of angiosperms, especially herbaceous species, remain debated.East Eurasia.Angiosperms.We compiled distributions of 43,023 angiosperm species in east Eurasia at spatial resolution (100 × 100 km2). Using this newly compiled database and a species‐level phylogeny, we estimated the phylogenetic structure patterns for species with different growth forms. We explored the relationships of these patterns with contemporary environment and historical climate change to test predictions of the tropical niche conservatism hypothesis (TCH) and to compare the contribution of different hypotheses using generalised linear models and hierarchical partitioning.We found that phylogenetic structure of angiosperms displayed significant latitudinal gradients. Notably, phylogenetic structure patterns and their drivers differed between woody and herbaceous species. Actual evapotranspiration was the best predictor of phylogenetic structure patterns for all and herbaceous species, while the mean temperature of the coldest quarter was the best predictor for woody species. The effect of historical climate change on phylogenetic structure patterns was weak.Our results suggest that the TCH only explains the phylogenetic structure pattern of woody species, not herbaceous species. Moreover, contemporary climate influences the phylogenetic structure of angiosperms in east Eurasian by affecting herbaceous and woody species differently. [ABSTRACT FROM AUTHOR]

This study investigates the effect of a bio-surfactant gelatin-modified carbon nanotubes (g-CNTs) on the fiber interfacial property and matrix performance of carbon fiber-reinforced polymer (CFRP) composite. Transverse fiber bundle test (TFBT) and in situ three-point bending test were conducted to analyze the fiber/matrix interfacial normal strength (IFNS) and bulk mechanical performance of the CNTs–CFRP composite. The results showed that g-CNTs have superb affinity and uniformity wrapping on the surface of carbon fiber via 2 min electrophoretic deposition (EPD) under a concentration of 0.1 mg/mL and a voltage strength of 10 V/cm, resulting in an increase of 40.3% of IFNS and 22.1%/25.3% of flexural strength/modulus of CFRP composites. Meanwhile, g-CNTs can also evenly distribute in the resin matrix with an improvement of 12.6% of IFNS and 20.3%/11.4% of flexural strength/modulus of CFRP composites under 0.1 wt% loading. This study provides a mechanism basis for the subsequent introduction of g-CNTs for the development of advanced CNT-reinforced CFRP composite.