•Hyperparameters are optimally selected by utilising random search methodology based on highly efficient implementation.•Handling missing data by using the mean of the available data significantly ...increase predictive performance.•The trained models based on GBR and XGBoost perform better than those of SVR and MLP.•Significant improvement of the current approach in terms of both prediction accuracy and computational effort.
In this study, an efficient implementation of machine learning models to predict compressive and tensile strengths of high-performance concrete (HPC) is presented. Four predictive algorithms including support vector regression (SVR), multilayer perceptron (MLP), gradient boosting regressor (GBR), and extreme gradient boosting (XGBoost) are employed. The process of hyperparameter tuning is based on random search that results in trained models with better predictive performances. In addition, the missing data is handled by filling with the mean of the available data which allows more information to be used in the training process. The results on two popular datasets of compressive and tensile strengths of high performance concrete show significant improvement of the current approach in terms of both prediction accuracy and computational effort. The comparative studies reveal that, for this particular prediction problem, the trained models based on GBR and XGBoost perform better than those of SVR and MLP.
•Developing methods for a late fusion of multiple Deep Neural Network models for better performance.•Proposing methods to determine priority weights for models.•Comparison between methods to ...determine optimal solution.•Allow to classify in near real-time.
In agriculture, many types of weeds have a harmful impact on agricultural productivity. Recognizing weeds and understanding the threat they pose to farmlands is a significant challenge because many weeds are quite similar in their external structure, making it difficult to classify them. A weeds classification approach with high accuracy and quick processing should be incorporated into automatic devices in smart agricultural systems to solve this problem. In this study, we develop a novel classification approach via a voting method by using the late fusion of multimodal Deep Neural Networks (DNNs). The score vector used for voting is calculated by either using Bayesian conditional probability-based method or by determining priority weights so that better DNNs models have a higher contribution to scoring. We experimentally studied the Plant Seedlings and Chonnam National University (CNU) Weeds datasets with 5 DNN models: NASNet, Resnet, Inception–Resnet, Mobilenet, and VGG. The results show that our methods achieved an accuracy of 97.31% on the Plant Seedlings dataset, and 98.77% accuracy on the CNU Weeds dataset. Furthermore, our framework can classify an image in near real-time.
Composting is a promising technology to decompose organic waste into humus-like high-quality compost, which can be used as organic fertilizer. However, greenhouse gases (N2O, CO2, CH4) and odorous ...emissions (H2S, NH3) are major concerns as secondary pollutants, which may pose adverse environmental and health effects. During the composting process, nitrogen cycle plays an important role to the compost quality. This review aimed to (1) summarizes the nitrogen cycle of the composting, (2) examine the operational parameters, microbial activities, functions of enzymes and genes affecting the nitrogen cycle, and (3) discuss mitigation strategies for nitrogen loss. Operational parameters such as moisture, oxygen content, temperature, C/N ratio and pH play an essential role in the nitrogen cycle, and adjusting them is the most straightforward method to reduce nitrogen loss. Also, nitrification and denitrification are the most crucial processes of the nitrogen cycle, which strongly affect microbial community dynamics. The ammonia-oxidizing bacteria or archaea (AOB/AOA) and the nitrite-oxidizing bacteria (NOB), and heterotrophic and autotrophic denitrifiers play a vital role in nitrification and denitrification with the involvement of ammonia monooxygenase (amoA) gene, nitrate reductase genes (narG), and nitrous oxide reductase (nosZ). Furthermore, adding additives such as struvite salts (MgNH4PO4·6H2O), biochar, and zeolites (clinoptilolite), and microbial inoculation, namely Bacillus cereus (ammonium strain), Pseudomonas donghuensis (nitrite strain), and Bacillus licheniformis (nitrogen fixer) can help control nitrogen loss. This review summarized critical issues of the nitrogen cycle and nitrogen loss in order to help future composting research with regard to compost quality and air pollution/odor control.
Display omitted
•The nitrogen cycle is the crucial biochemical process of organic waste composting.•Nitrification and denitrification are the main steps of the nitrogen cycle.•Proteobacteria and Thaumarchaeota are dominant microorganisms in thermophilic phase.•Nitrogen loss mainly occurs via volatilization and leaching during the composting.•Adding biochar, zeolite, and microbial incubation can mitigate nitrogen loss.
Lichens produce secondary metabolites that have many pharmaceutical activities such as antimicrobial, antioxidant, antiviral, anticancer, antigenotoxic, anti-inflammatory, analgesic and antipyretic ...activities. However, there is limited research on their efflux pump inhibitory activities. Twelve phytochemicals were isolated from
, and their activity of AcrAB-TolC efflux pump inhibition was evaluated. Four potential compounds, which are diffractaic acid
, 8' -
- methylstictic acid
, 3-hydroxy-4-(methoxycarbonyl)-2,5-dimethylphenyl 2,4-dimethoxy-3,6-dimethylbenzoate (
and 3-hydroxy-4-(methoxycarbonyl)-2,5-dimethylphenyl 2-hydroxy-4-methoxy-3,6-dimethylbenzoate
, were found by virtual screening using pharmacophore and 2D-QSAR model. Compound
exhibited AcrB inhibition activity
with an accumulation H33342 percentage compared with untreated control of 202% at a concentration of 50 µM and increased the antibacterial activity of levofloxacin by four-fold at a concentration of 200 µM. By molecular docking and molecular dynamics (MD) simulation, the binding affinity of depside and depsidone derivatives to AcrB was also clarified. Despite the poor docking score to the AcrB binding site, compound
was the most stable among the four complexes at 20 ns of MD simulation. The analysis of long MD at 100 ns indicated that compound
interacts strongly with the residues in the distal pocket, creating a stable complex with
of
kcal.mol-1. According to the ADMETlab 2.0 web server's predictions of pharmacokinetics and toxicities, compound
has the potential for drug development.Communicated by Ramaswamy H. Sarma.
In recent years, with tremendous progresses of deep learning in multiple disciplines, there are several advanced sequential neural-network (NN) based architectures (e.g., recurrent neural ...network-RNN, Auto-Encoding-AE, transformer, etc.) have been proposed. Recently, there are several well-known GNN-based architectures like as graph convolutional network (GNN) have been proposed to deal with challenges related to the global representation preservation of text. However, most of recent proposed GNN-based text-embedding models still be unable to integrate the global structure with the semantic sequential representations of words/sentences into the unified textual embedding space. Moreover, they are also considered as unable to learn the rich context-varied representations of words. In order to tackle aforementioned challenges, in this paper we proposed a novel integrated text graph representation learning approach, named as: GOWSeqGCN. Our proposed GOWSeqGCN is an integrated semantic graph-of-words sequential textual representation under the graph convolutional network framework. In order to demonstrate for the effectiveness of our proposed GOWSeqGCN model in comparing with recent state-of-the-art text representation learning baselines, we conducted extensive experiments in benchmark textual datasets. The experimental outputs showed the outperformances and necessary of our proposed ideas in this paper.
Resource recovery in wastewater treatment is the major challenge for achieving sustainable development. Further than only treatment, the recovery and reuse of nitrogen (N) and phosphorus (P) as a ...slow release fertilizer from actual swine wastewater (SW) was performed in this study. The development of a novel technique of converting soft sludge into hard particles using homogeneous granulation process in a fluidized bed reactor helps address the recovery. Under conditions of pH 8.5 ± 0.2, an Mg:P ratio of 1.2, total suspended solids (TSS) < 100 mg/L, and a 25 m/h up-flow velocity, the removal of N and P reached 69.8% and 96.8%, respectively. XRD and SEM-EDS analysis confirmed that N and P were crystallized as magnesium ammonium phosphate hexahydrate (struvite, NH4MgPO4·6H2O) phase with a purity greater than 68.3% and an average particle size of 0.88 mm. The increase in total suspended solids resulted in not only decreases in the removal efficiency and struvite purity but also an irregular morphology in the struvite crystals. Furthermore, increasing the magnesium concentration significantly reduced the crystal size of the pellets. The optimal up-flow velocity should be 1.5–2.4 times the minimum fluidization velocity (12.75 m/h). A cost-benefit estimation of actual swine wastewater treatment indicated a profit of approximately 0.345 USD per m3-actual swine wastewater. Fluidized-bed homogeneous granulation was proven to be a promising technology for recovering N and P from swine wastewater.
Display omitted
•Struvite (diameter = 0.88 mm) was recovered from actual swine and Mg-rich wastewater.•The maximum recoveries were 69.8% and 96.8% for N and P, respectively.•Total suspended solids greatly influenced N and P removal as well as struvite size and purity.•The purity of the struvite pellets was affected by solution pH, total suspended solids and Mg concentration.•Homogeneous granulation was the main mechanism for N and P removal.
In considering the problem of saving spaces during the transportation of items from one station to another, for example, in warehouses, factories, hospitals, etc., an automatic transportation system ...(ATS) that could take advantage of the above ceiling spaces for the transportation of products is considered. Such a system guarantees that the activities occurring in the floor area will be maintained as usual. To achieve this requirement, the ceiling spaces of a building are used to construct an automatic multi-station multi-container (MSMC) transportation system. This system can transport items from one place to another in the whole system. This system is designed to utilize the spaces above the ceiling, and it has the advantage of saving floor space for transportation operations. This will increase the operational capability of the industries and also improve the productivity of the industry in which this system is implemented. The entire transportation system includes (1) the essential conveying system (which is a functional conveyor module with a specified number of containers); (2) the control block that can monitor and operate the system; and (3) the sensor block for detecting and identifying the containers. The content of this article focuses on the introduction of the mechanical system (1); the control system (2); and the operating principle of the whole system (3).
Capsule Network with Shortcut Routing DANG, Thanh Vu; VO, Hoang Trong; YU, Gwang Hyun ...
IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences,
08/2021, Letnik:
E104.A, Številka:
8
Journal Article
Recenzirano
Odprti dostop
Capsules are fundamental informative units that are introduced into capsule networks to manipulate the hierarchical presentation of patterns. The part-hole relationship of an entity is learned ...through capsule layers, using a routing-by-agreement mechanism that is approximated by a voting procedure. Nevertheless, existing routing methods are computationally inefficient. We address this issue by proposing a novel routing mechanism, namely “shortcut routing”, that directly learns to activate global capsules from local capsules. In our method, the number of operations in the routing procedure is reduced by omitting the capsules in intermediate layers, resulting in lighter routing. To further address the computational problem, we investigate an attention-based approach, and propose fuzzy coefficients, which have been found to be efficient than mixture coefficients from EM routing. Our method achieves on-par classification results on the Mnist (99.52%), smallnorb (93.91%), and affNist (89.02%) datasets. Compared to EM routing, our fuzzy-based and attention-based routing methods attain reductions of 1.42 and 2.5 in terms of the number of calculations.
There is scarce information regarding hospital-acquired infections (HAIs) among children in resource-constrained settings. This study aims to measure prevalence of HAIs in Vietnamese pediatric ...hospitals.Monthly point prevalence surveys (PPSs) in 6 pediatric intensive care units (ICUs) in 3 referral hospitals during 1 year.A total of 1363 cases (1143 children) were surveyed, 59.9% male, average age 11 months. Admission sources were: other hospital 49.3%, current hospital 36.5%, and community 15.3%. Reasons for admission were: infectious disease (66%), noninfectious (20.8%), and surgery/trauma (11.3%). Intubation rate was 47.8%, central venous catheter 29.4%, peripheral venous catheter 86.2%, urinary catheter 14.6%, and hemodialysis/filtration 1.7%. HAI was diagnosed in 33.1% of the cases: pneumonia (52.2%), septicemia (26.4%), surgical site infection (2%), and necrotizing enterocolitis (2%). Significant risk factors for HAI included age under 7 months, intubation and infection at admission. Microbiological findings were reported in 212 cases (43%) with 276 isolates: 50 Klebsiella pneumoniae, 46 Pseudomonas aeruginosa, and 39 Acinetobacter baumannii, with carbapenem resistance detected in 55%, 71%, and 65%, respectively. Staphylococcus aureus was cultured in 18 cases, with 81% methicillin-resistant Staphylococcus aureus. Most children (87.6%) received antibiotics, with an average of 1.6 antibiotics per case. Colistin was administered to 96 patients, 93% with HAI and 49% with culture confirmed carbapenem resistance.The high prevalence of HAI with carbapenem resistant gram-negative strains and common treatment with broad-spectrum antibiotics and colistin suggests that interventions are needed to prevent HAI and to optimize antibiotic use.