Shandong Jianzhu University

OBJECTIVE: Vast 12-lead ECGs repositories provide opportunities to develop new machine learning approaches for creating accurate and automatic diagnostic systems for cardiac abnormalities. However, most 12-lead ECG classification studies are trained, tested, or developed in single, small, or relatively homogeneous datasets. In addition, most algorithms focus on identifying small numbers of cardiac arrhythmias that do not represent the complexity and difficulty of ECG interpretation. This work addresses these issues by providing a standard, multi-institutional database and a novel scoring metric through a public competition: the PhysioNet/Computing in Cardiology Challenge 2020. APPROACH: A total of 66 361 12-lead ECG recordings were sourced from six hospital systems from four countries across three continents; 43 101 recordings were posted publicly with a focus on 27 diagnoses. For the first time in a public competition, we required teams to publish open-source code for both training and testing their algorithms, ensuring full scientific reproducibility. MAIN RESULTS: A total of 217 teams submitted 1395 algorithms during the Challenge, representing a diversity of approaches for identifying cardiac abnormalities from both academia and industry. As with previous Challenges, high-performing algorithms exhibited significant drops ([Formula: see text]10%) in performance on the hidden test data. SIGNIFICANCE: Data from diverse institutions allowed us to assess algorithmic generalizability. A novel evaluation metric considered different misclassification errors for different cardiac abnormalities, capturing the outcomes and risks of different diagnoses. Requiring both trained models and code for training models improved the generalizability of submissions, setting a new bar in reproducibility for public data science competitions.

Traditional polyamide-based interfacial polymerized nanofiltration (NF) membranes exhibit upper bound features between water permeance and salt selectivity. Breaking the limits of the permeability and rejections of these composite NF membranes are highly desirable for water desalination. Herein, a high-performance NF membrane (TFC-P) was fabricated via interfacial polymerization on the poly(vinyl alcohol) (PVA) interlayered poly(ether sulfone) (PES) ultrafiltration support. Owing to the large surface area, great hydrophilicity, and high porosity of the PES–PVA support, a highly cross-linked polyamide separating layer was formed with a thickness of 9.6 nm, which was almost 90% thinner than that of the control membrane (TFC-C). In addition, the TFC-P possessed lower ζ-potential, smaller pore size, and greater surface area compared to that of the TFC-C, achieving an ultrahigh water permeance of 31.4 L m–2 h–1 bar–1 and a 99.4% Na2SO4 rejection. Importantly, the PVA interlayer strategy was further applied to a pilot NF production line and the fabricated membranes presented stable water flux and salt rejections as comparable to the lab-scaled membranes. The outstanding properties of the PVA-interlayered NF membranes highlight the feasibility of the fabrication method for practical applications, which provides a new avenue to develop robust polyamide-based NF desalination membranes for environmental water treatment.

Nanofiltration (NF) membranes have been widely applied in many important environmental applications, including water softening, surface/groundwater purification, wastewater treatment, and water reuse. In recent years, a new class of piperazine (PIP)-based NF membranes featuring a crumpled polyamide layer has received considerable attention because of their great potential for achieving dramatic improvements in membrane separation performance. Since the report of novel crumpled Turing structures that exhibited an order of magnitude enhancement in water permeance ( Science 2018, 360 (6388), 518−521), the number of published research papers on this emerging topic has grown exponentially to approximately 200. In this critical review, we provide a systematic framework to classify the crumpled NF morphologies. The fundamental mechanisms and fabrication methods involved in the formation of these crumpled morphologies are summarized. We then discuss the transport of water and solutes in crumpled NF membranes and how these transport phenomena could simultaneously improve membrane water permeance, selectivity, and antifouling performance. The environmental applications of these emerging NF membranes are highlighted, and future research opportunities/needs are identified. The fundamental insights in this review provide critical guidance on the further development of high-performance NF membranes tailored for a wide range of environmental applications.

As a universal process in multicellular organisms, including animals and plants, cells usually emit danger signals when suffering from attacks of microbes and herbivores, or physical damage. These signals, termed as damage-associated molecular patterns (DAMPs), mainly include cell wall or extracellular protein fragments, peptides, nucleotides, and amino acids. Once exposed on cell surfaces, DAMPs are detected by plasma membrane-localized receptors of surrounding cells to regulate immune responses against the invading organisms and promote damage repair. DAMPs may also act as long-distance mobile signals to mediate systemic wounding responses. Generation, release, and perception of DAMPs, and signaling events downstream of DAMP perception are all rigorously modulated by plants. These processes integrate together to determine intricate mechanisms of DAMP-triggered immunity in plants. In this review, we present an extensive overview on our current understanding of DAMPs in plant immune system.

Building energy consumption prediction plays an important role in improving the energy utilization rate through helping building managers to make better decisions. However, as a result of randomness and noisy disturbance, it is not an easy task to realize accurate prediction of the building energy consumption. In order to obtain better building energy consumption prediction accuracy, an extreme deep learning approach is presented in this paper. The proposed approach combines stacked autoencoders (SAEs) with the extreme learning machine (ELM) to take advantage of their respective characteristics. In this proposed approach, the SAE is used to extract the building energy consumption features, while the ELM is utilized as a predictor to obtain accurate prediction results. To determine the input variables of the extreme deep learning model, the partial autocorrelation analysis method is adopted. Additionally, in order to examine the performances of the proposed approach, it is compared with some popular machine learning methods, such as the backward propagation neural network (BPNN), support vector regression (SVR), the generalized radial basis function neural network (GRBFNN) and multiple linear regression (MLR). Experimental results demonstrate that the proposed method has the best prediction performance in different cases of the building energy consumption.

Road defect detection is a crucial task for promptly repairing road damage and ensuring road safety. Traditional manual detection methods are inefficient and costly. To overcome this issue, we propose an enhanced road defect detection algorithm called BL-YOLOv8, which is based on YOLOv8s. In this study, we optimized the YOLOv8s model by reconstructing its neck structure through the integration of the BiFPN concept. This optimization reduces the model's parameters, computational load, and overall size. Furthermore, to enhance the model's operation, we optimized the feature pyramid layer by introducing the SimSPPF module, which improves its speed. Moreover, we introduced LSK-attention, a dynamic large convolutional kernel attention mechanism, to expand the model's receptive field and enhance the accuracy of object detection. Finally, we compared the enhanced YOLOv8 model with other existing models to validate the effectiveness of our proposed improvements. The experimental results confirmed the effective recognition of road defects by the improved YOLOv8 algorithm. In comparison to the original model, an improvement of 3.3% in average precision mAP@0.5 was observed. Moreover, a reduction of 29.92% in parameter volume and a decrease of 11.45% in computational load were achieved. This proposed approach can serve as a valuable reference for the development of automatic road defect detection methods.

In plants, innate immune responses are initiated by plasma membrane-located pattern recognition receptors (PRRs) upon recognition of elicitors, including exogenous pathogen-associated molecular patterns (PAMPs) and endogenous damage-associated molecular patterns (DAMPs). Arabidopsis thaliana produces more than 1000 secreted peptide candidates, but it has yet to be established whether any of these act as elicitors. Here we identified an A. thaliana gene family encoding precursors of PAMP-induced secreted peptides (prePIPs) through an in-silico approach. The expression of some members of the family, including prePIP1 and prePIP2, is induced by a variety of pathogens and elicitors. Subcellular localization and proteolytic processing analyses demonstrated that the prePIP1 product is secreted into extracellular spaces where it is cleaved at the C-terminus. Overexpression of prePIP1 and prePIP2, or exogenous application of PIP1 and PIP2 synthetic peptides corresponding to the C-terminal conserved regions in prePIP1 and prePIP2, enhanced immune responses and pathogen resistance in A. thaliana. Genetic and biochemical analyses suggested that the receptor-like kinase 7 (RLK7) functions as a receptor of PIP1. Once perceived by RLK7, PIP1 initiates overlapping and distinct immune signaling responses together with the DAMP PEP1. PIP1 and PEP1 cooperate in amplifying the immune responses triggered by the PAMP flagellin. Collectively, these studies provide significant insights into immune modulation by Arabidopsis endogenous secreted peptides.

This work investigates a finite-time adaptive fuzzy tracking control problem for a class of nonstrict feedback nonlinear systems from a new point of view. A new concept, named finite-time performance function (FTPF), is defined in this paper for the first time. Moreover, a finite-time adaptive state feedback fuzzy tracking controller is derived based on fuzzy approximation, backstepping technique and prescribed performance control (PPC), which guarantees that all the signals of the closed-loop system are bounded, the output tracking error converges to a prescribed arbitrarily small region within a finite-time interval, and maximum overshoot is not more than a predefined level. In addition, a controller design process is given which is less complex than the existing finite-time control design methods. Three simulation studies are provided to verify the feasibility and effectiveness of the theoretical finding in this study.

Molecular dynamics provides a powerful tool to understand the elusive structure–performance relationship of asphalts. The combined molecular models were selected to investigate the interactions of the saturate, aromatic, resin, and asphaltene (SARA) four fractions and the correlation between fractions and the “bee-like structures” by atomic force microscopy in asphalts. The results showed that van der Waals was the main force to control intermolecular interactions. The arrangement of SARA fractions largely conformed to the modern colloid theory. However, some alkanes, sulfides, and condensed aromatics had different behaviors. Long-chain alkanes inserted into layers of asphaltenes, and small sulfides without long alkyl chains adhered to large sulfides or asphaltenes; nevertheless, counterpart condensed aromatics became much closer to those molecules. Strong interactions between the dispersed phase and continuous phase generated a larger size and greater number of “bee structures”. Asphaltenes played as a core, and long-chain paraffin played as an inducer, to affect the distribution of “bee structures”.

International audience

In this technical note, constructive control techniques have been proposed for controlling feedforward nonlinear time-delay systems. The nonlinear terms admit an incremental rate depending on the input or delayed input. Based on the Lyapunov-Razumikhin theorem and Lyapunov-Krasovskii theorem, the delay-independent feedback controllers are explicitly constructed such that the closed-loop systems are globally asymptotically stable. An example is given to demonstrate the effectiveness of the proposed design procedure.

Knowledge economy era is an era driven by innovation, mainly based on the input of intangible assets which plays decisive roles in the long-term development of enterprises. The product value of enterprises is largely determined by their intellectual capital. Therefore, as pillars of China’s economy, construction enterprises must strengthen their investments in intellectual capital, and to achieve competitiveness in the market, enterprises must share knowledge with the other members of their networks. This study explores the relationship among the intellectual capital, knowledge sharing, and innovation performance of construction enterprises and the mediating effect of knowledge sharing on the relationship between intellectual capital and innovation performance by using data collected from a questionnaire survey. These data are analyzed along with the aforementioned relationships by using SPSS and a structural equation model. The findings indicate that intellectual capital not only has a direct positive influence on the innovation performance of construction enterprises but also positively affects their innovation performance through knowledge sharing. This paper concludes by presenting its limitations and the implications of its findings.

This paper proposed an improved Grey Wolf Optimizer (GWO) to resolve the problem of instability and convergence accuracy when GWO is used as a meta-heuristic algorithm with strong optimal search capability in the path planning for mobile robots. We improved chaotic tent mapping to initialize the wolves to enhance the global search ability and used a nonlinear convergence factor based on the Gaussian distribution change curve to balance the global and local searchability. In addition, an improved dynamic proportional weighting strategy is proposed that can update the positions of grey wolves so that the convergence of this algorithm can be accelerated. The proposed improved GWO algorithm results are compared with the other eight algorithms through several benchmark function test experiments and path planning experiments. The experimental results show that the improved GWO has higher accuracy and faster convergence speed.

The performance of Cd, Ni, and Cu release from river sediment at different pH was investigated by a leaching test using deionised water and river water as leachants. Visual MINTEQ geochemical software was used to model the experimental results to predict heavy metal release from sediments. The distribution and speciation of heavy metals in the sediments after leaching test were analyzed by Tessier sequential extraction. Leaching test results showed that the release amounts of Cd, Ni, and Cu are in the range of 10.2–27.3 mg·kg −1 , 80.5–140.1 mg·kg −1 , and 6.1–30.8 mg·kg −1 , respectively, with deionised water as leachant at different pH. As far as the river water was used as the leaching solution in the test, the results show similar metal leaching contents and tendencies to that of the deionised water as leaching solution. The results of Tessier sequential extraction indicate that Cd of residual fraction easily forms obvious precipitate under the acidic condition, especially in the range of pH 0–4 with the residual of Cd over 50% of the total Cd in the sediment. The exchangeable content of Ni decreases with the increase of pH under the range of 0–5. The Fe-Mn oxide fraction of Cu in the sediments changes significantly from pH 0 to pH 9. Based on the effect of pH on the leaching of Cd, Ni, and Cu from the polluted sediment in the tests, more accurate information could be obtained to assess the risk related to metal release from sediments once it is exposed to the changed acid/alkali water conditions.

The Yangtze River Basin has a wide range and complicated topography. In recent years, under the background of climate and land cover change, the ecological response of the whole ecological quality of the Yangtze River Basin is still unknown. To reveal the spatiotemporal changes in ecological quality in the Yangtze River Basin from 2001 to 2019 and their relationship with environmental and topographical factors, this study used the Google Earth Engine (GEE) platform to calculate the remote sensing ecological index (RSEI) based on the Moderate-resolution Imaging Spectroradiometer (MODIS) product image set, combined with the digital elevation data set and statistical yearbook. The data evaluated the ecological quality of the Yangtze River Basin and analyzed its causes. The results showed that: 1) the average RSEI of the Yangtze River Basin showed an overall upward trend, the growth rate was 0.027 (year−1), and the variation ranged from 0.5 to 0.568. The overall ecological quality rank was mainly neutral and slightly good. 2) The ecological quality of 85.7% of the Yangtze River Basin remains stable. A total of 11.2% of the regional ecological quality is improving, and 3.1% of the regional ecological quality is declining. Areas with reduced ecological quality are concentrated in the Hengduan Mountains. The dominant LST factor drives the deterioration of its ecological quality at a rate of −1.06 (year−1). The areas with improved ecological quality are concentrated in the upper and middle reaches of the Yangtze River. The dominant WET factor drives its ecological quality to improve at a rate of 0.27 (year−1). 3) From the perspective of topography, the ecological quality of the Yangtze River basin shows a wave-like decline and first rises and then falls in elevation and slope (the elevation is bounded by 2000 m and 6000 m, and the slope is bounded by 15°.). The average RSEI of the Yangtze River Basin is the highest on the northwest slope (0.554), and the ecological quality of sunny slopes is generally higher than that of shady slopes. The research shows that from 2001 to 2019, the overall ecological quality of the Yangtze River Basin has improved and evolved, but the ecological quality of the Hengduan Mountains has declined. Therefore, implementing different ecological protection policies in different regions is an important strategy for enhancing the stability of the ecosystem.

Biodegradable active food packaging addresses key environmental issues including plastic waste and food waste.

• The novel analysis strategy developed to understand data normalization method. • The significant influence of data normalization on the predictive capabilities of various ANN models • More effective combinations of ANN models with specific data normalization strategies • Evaluating the correlation between each data normalization method on the energy consumption The study investigates the impact of data normalization on the prediction of electricity consumption in buildings using four multilayer Artificial Neural Networks (ANN) algorithms: Long Short-Term Memory Networks (LSTM), Levenberg-Marquardt Back-propagation (LMBP), Recurrent Neural Networks (RNN), and General Regression Neural Network (GRNN). Four data normalization approaches, Min-Max Scaling, Mean, Z-score, and Gaussian function were assessed on experimental datasets. The LSTM algorithm, when combined with Min-Max normalization, showed the most favorable predictive capabilities, with a low Coefficient of Variation of the Root Mean Square Error (CVRMSE) of 10.3 and Normalized Mean Bias Error (NMBE) of 0.6. The remaining three normalization approaches showed satisfactory concordance with empirical data, but with slight disparities in precision. The LMBP model, when using Z-score normalization, had favorable performance in forecasting electricity consumption, but the discrepancies across the models were not significant. The Recurrent Neural Network (RNN) model, when used with Gaussian normalization, exhibited the most favorable performance, with the lowest Coefficient of Variation of Root Mean Square Error (CVRMSE) at 11.8 and Normalized Mean Biased Error (NMBE) at 0.6. The Generalized Regression Neural Network (GRNN) model, trained on unprocessed data, exhibited superior performance, with the lowest Coefficient of Variation of Root Mean Square Error (CVRMSE) at 19.2 and NMBE at 1.0. In conclusion, the study highlights the significant influence of data normalization on the predictive capabilities of various ANN models, suggesting that careful use of data normalization techniques can significantly improve the accuracy of electricity consumption forecasting in buildings.

Image-text retrieval is a fundamental and crucial branch in information retrieval. Although much progress has been made in bridging vision and language, it remains challenging because of the difficult intra-modal reasoning and cross-modal alignment. Existing modality interaction methods have achieved impressive results on public datasets. However, they heavily rely on expert experience and empirical feedback towards the design of interaction patterns, therefore, lacking flexibility. To address these issues, we develop a novel modality interaction modeling network based upon the routing mechanism, which is the first unified and dynamic multimodal interaction framework towards image-text retrieval. In particular, we first design four types of cells as basic units to explore different levels of modality interactions, and then connect them in a dense strategy to construct a routing space. To endow the model with the capability of path decision, we integrate a dynamic router in each cell for pattern exploration. As the routers are conditioned on inputs, our model can dynamically learn different activated paths for different data. Extensive experiments on two benchmark datasets, i.e., Flickr30K and MS-COCO, verify the superiority of our model compared with several state-of-the-art baselines.

Flexible job shop scheduling problem (FJSP) has attracted research interests as it can significantly improve the energy, cost, and time efficiency of production. As one type of reinforcement learning, deep Q-network (DQN) has been applied to solve numerous realistic optimization problems. In this study, a DQN model is proposed to solve a multiobjective FJSP with crane transportation and setup times (FJSP-CS). Two objectives, i.e., makespan and total energy consumption, are optimized simultaneously based on weighting approach. To better reflect the problem realities, eight different crane transportation stages and three typical machine states including processing, setup, and standby are investigated. Considering the complexity of FJSP-CS, an identification rule is designed to organize the crane transportation in solution decoding. As for the DQN model, 12 state features and seven actions are designed to describe the features in the scheduling process. A novel structure is applied in the DQN topology, saving the calculation resources and improving the performance. In DQN training, double deep Q-network technique and soft target weight update strategy are used. In addition, three reported improvement strategies are adopted to enhance the solution qualities by adjusting scheduling assignments. Extensive computational tests and comparisons demonstrate the effectiveness and advantages of the proposed method in solving FJSP-CS, where the DQN can choose appropriate dispatching rules at various scheduling situations.

BACKGROUND: Accurate and meaningful dose metrics are a basic requirement for in vitro screening to assess potential health risks of engineered nanomaterials (ENMs). Correctly and consistently quantifying what cells "see," during an in vitro exposure requires standardized preparation of stable ENM suspensions, accurate characterizatoin of agglomerate sizes and effective densities, and predictive modeling of mass transport. Earlier transport models provided a marked improvement over administered concentration or total mass, but included assumptions that could produce sizable inaccuracies, most notably that all particles at the bottom of the well are adsorbed or taken up by cells, which would drive transport downward, resulting in overestimation of deposition. METHODS: Here we present development, validation and results of two robust computational transport models. Both three-dimensional computational fluid dynamics (CFD) and a newly-developed one-dimensional Distorted Grid (DG) model were used to estimate delivered dose metrics for industry-relevant metal oxide ENMs suspended in culture media. Both models allow simultaneous modeling of full size distributions for polydisperse ENM suspensions, and provide deposition metrics as well as concentration metrics over the extent of the well. The DG model also emulates the biokinetics at the particle-cell interface using a Langmuir isotherm, governed by a user-defined dissociation constant, K(D), and allows modeling of ENM dissolution over time. RESULTS: Dose metrics predicted by the two models were in remarkably close agreement. The DG model was also validated by quantitative analysis of flash-frozen, cryosectioned columns of ENM suspensions. Results of simulations based on agglomerate size distributions differed substantially from those obtained using mean sizes. The effect of cellular adsorption on delivered dose was negligible for K(D) values consistent with non-specific binding (> 1 nM), whereas smaller values (≤ 1 nM) typical of specific high-affinity binding resulted in faster and eventual complete deposition of material. CONCLUSIONS: The advanced models presented provide practical and robust tools for obtaining accurate dose metrics and concentration profiles across the well, for high-throughput screening of ENMs. The DG model allows rapid modeling that accommodates polydispersity, dissolution, and adsorption. Result of adsorption studies suggest that a reflective lower boundary condition is appropriate for modeling most in vitro ENM exposures.