In this study, we used unidirectional and bidirectional long short-term memory (LSTM) deep learning networks for Chinese news classification and characterized the effects of contextual information on ...text classification, achieving a high level of accuracy. A Chinese glossary was created using jieba-a word segmentation tool-stop-word removal, and word frequency analysis. Next, word2vec was used to map the processed words into word vectors, creating a convenient lookup table for word vectors that could be used as feature inputs for the LSTM model. A bidirectional LSTM (BiLSTM) network was used for feature extraction from word vectors to facilitate the transfer of information in both the backward and forward directions to the hidden layer. Subsequently, an LSTM network was used to perform feature integration on all the outputs of the BiLSTM network, with the output from the last layer of the LSTM being treated as the mapping of the text into a feature vector. The output feature vectors were then connected to a fully connected layer to construct a feature classifier using the integrated features, finally classifying the news articles. The hyperparameters of the model were optimized based on the loss between the true and predicted values using the adaptive moment estimation (Adam) optimizer. Additionally, multiple dropout layers were added to the model to reduce overfitting. As text classification models for Chinese news articles, the Bi-LSTM and unidirectional LSTM models obtained f1-scores of 94.15% and 93.16%, respectively, with the former outperforming the latter in terms of feature extraction.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Although maternal deaths are rare in developed regions, the morbidity associated with severe postpartum hemorrhage (SPPH) remains a major problem. To determine the prevalence and risk factors of ...SPPH, we analyzed data of women who gave birth in Guangzhou Medical Centre for Critical Pregnant Women, which received a large quantity of critically ill obstetric patients who were transferred from other hospitals in Southern China.
In this study, we conducted a retrospective case-control study to determine the prevalence and risk factors for SPPH among a cohort of women who gave birth after 28 weeks of gestation between January 2015 and August 2019. SPPH was defined as an estimated blood loss ≥1000 mL and total blood transfusion≥4 units. Logistic regression analysis was used to identify independent risk factors for SPPH.
SPPH was observed in 532 mothers (1.56%) among the total population of 34,178 mothers. Placenta-related problems (55.83%) were the major identified causes of SPPH, while uterine atony without associated retention of placental tissues accounted for 38.91%. The risk factors for SPPH were maternal age < 18 years (adjusted OR aOR = 11.52, 95% CI: 1.51-87.62), previous cesarean section (aOR = 2.57, 95% CI: 1.90-3.47), history of postpartum hemorrhage (aOR = 4.94, 95% CI: 2.63-9.29), conception through in vitro fertilization (aOR = 1.78, 95% CI: 1.31-2.43), pre-delivery anemia (aOR = 2.37, 95% CI: 1.88-3.00), stillbirth (aOR = 2.61, 95% CI: 1.02-6.69), prolonged labor (aOR = 5.24, 95% CI: 3.10-8.86), placenta previa (aOR = 9.75, 95% CI: 7.45-12.75), placenta abruption (aOR = 3.85, 95% CI: 1.91-7.76), placenta accrete spectrum (aOR = 8.00, 95% CI: 6.20-10.33), and macrosomia (aOR = 2.30, 95% CI: 1.38-3.83).
Maternal age < 18 years, previous cesarean section, history of PPH, conception through IVF, pre-delivery anemia, stillbirth, prolonged labor, placenta previa, placental abruption, PAS, and macrosomia were risk factors for SPPH. Extra vigilance during the antenatal and peripartum periods is needed to identify women who have risk factors and enable early intervention to prevent SPPH.
A resilience-oriented service restoration method using microgrids to restore critical load after natural disasters is proposed in this paper. Considering the scarcity of power generation resources, ...the concept of continuous operating time (COT) is introduced to determine the availability of microgrids for critical load restoration and to assess the service time. Uncertainties induced by intermittent energy sources and load are also taken into account. The critical load restoration problem is modeled as a chance-constrained stochastic program. A Markov chain-based operation model is designed to describe the stochastic energy variations within microgrids, based on which the COT is assessed. A two-stage heuristic is developed for the critical load restoration problem. First, a strategy table containing the information of all feasible restoration paths is established. Then the critical load restoration strategy is obtained by solving a linear integer program. Numerical simulations are performed on the IEEE 123-node feeder system under several scenarios to demonstrate the effectiveness of the proposed method. The impacts of fault locations, available generation resources, and load priority on the restoration strategy are discussed.
Polymorphism in the apolipoprotein E (APOE) gene is a major genetic risk determinant of late-onset Alzheimer disease (AD), with the APOE*ε4 allele conferring an increased risk and the APOE*ε2 allele ...conferring a decreased risk relative to the common APOE*ε3 allele. Strong evidence from clinical and basic research suggests that a major pathway by which APOE4 increases the risk of AD is by driving earlier and more abundant amyloid pathology in the brains of APOE*ε4 carriers. The number of amyloid-β (Aβ)-dependent and Aβ-independent pathways that are known to be differentially modulated by APOE isoforms is increasing. For example, evidence is accumulating that APOE influences tau pathology, tau-mediated neurodegeneration and microglial responses to AD-related pathologies. In addition, APOE4 is either pathogenic or shows reduced efficiency in multiple brain homeostatic pathways, including lipid transport, synaptic integrity and plasticity, glucose metabolism and cerebrovascular function. Here, we review the recent progress in clinical and basic research into the role of APOE in AD pathogenesis. We also discuss how APOE can be targeted for AD therapy using a precision medicine approach.
Dynamics of shear Alfven waves and energetic particles are crucial to the performance of burning fusion plasmas. This article reviews linear as well as nonlinear physics of shear Alfven waves and ...their self-consistent interaction with energetic particles in tokamak fusion devices. More specifically, the review on the linear physics deals with wave spectral properties and collective excitations by energetic particles via wave-particle resonances. The nonlinear physics deals with nonlinear wave-wave interactions as well as nonlinear wave-energetic particle interactions. Both linear as well as nonlinear physics demonstrate the qualitatively important roles played by realistic equilibrium nonuniformities, magnetic field geometries, and the specific radial mode structures in determining the instability evolution, saturation, and, ultimately, energetic-particle transport. These topics are presented within a single unified theoretical framework, where experimental observations and numerical simulation results are referred to elucidate concepts and physics processes.
To overcome devices' limitations in performing computation-intense applications, mobile edge computing (MEC) enables users to offload tasks to proximal MEC servers for faster task computation. ...However, the current MEC system design is based on average-based metrics, which fails to account for the ultra-reliable low-latency requirements in mission-critical applications. To tackle this, this paper proposes a new system design, where probabilistic and statistical constraints are imposed on task queue lengths, by applying extreme value theory . The aim is to minimize users' power consumption while trading off the allocated resources for local computation and task offloading. Due to wireless channel dynamics, users are reassociated to MEC servers in order to offload tasks using higher rates or accessing proximal servers. In this regard, a user-server association policy is proposed, taking into account the channel quality as well as the servers' computation capabilities and workloads. By marrying tools from Lyapunov optimization and matching theory, a two-timescale mechanism is proposed, where a user-server association is solved in the long timescale, while a dynamic task offloading and resource allocation policy are executed in the short timescale. The simulation results corroborate the effectiveness of the proposed approach by guaranteeing highly reliable task computation and lower delay performance, compared to several baselines.
Deep neural networks (DNN) have achieved unprecedented success in numerous machine learning tasks in various domains. However, the existence of adversarial examples raises our concerns in adopting ...deep learning to safety-critical applications. As a result, we have witnessed increasing interests in studying attack and defense mechanisms for DNN models on different data types, such as images, graphs and text. Thus, it is necessary to provide a systematic and comprehensive overview of the main threats of attacks and the success of corresponding countermeasures. In this survey, we review the state of the art algorithms for generating adversarial examples and the countermeasures against adversarial examples, for three most popular data types, including images, graphs and text.
In comparison to severe acute respiratory syndrome coronavirus (SARS-CoV), SARS-CoV-2 appears to be more contagious 1, and coronavirus disease 2019 (COVID-19) patients demonstrate varied clinical ...manifestations distinct from those seen in patients with SARS-CoV and Middle East respiratory syndrome coronavirus infections 2. Collective results from the clinical and epidemiological observations suggest a distinct viral–host interaction in COVID-19 patients. Profiling of the antibody response during SARS-CoV-2 infection may help improve our understanding of the viral–host interaction and the immunopathological mechanisms of the disease.
Humoral immune response to SARS-CoV-2 showed an early response of IgA, instead of IgM, in COVID-19 patients. As highlighted by this study, enhanced IgA responses observed in severe COVID-19 might confer damaging effects in severe COVID-19.
https://bit.ly/3fA7c1I
Vulnerability assessment is a requirement of NERC's cybersecurity standards for electric power systems. The purpose is to study the impact of a cyber attack on supervisory control and data ...acquisition (SCADA) systems. Compliance of the requirement to meet the standard has become increasingly challenging as the system becomes more dispersed in wide areas. Interdependencies between computer communication system and the physical infrastructure also become more complex as information technologies are further integrated into devices and networks. This paper proposes a vulnerability assessment framework to systematically evaluate the vulnerabilities of SCADA systems at three levels: system , scenarios , and access points . The proposed method is based on cyber systems embedded with the firewall and password models, the primary mode of protection in the power industry today. The impact of a potential electronic intrusion is evaluated by its potential loss of load in the power system. This capability is enabled by integration of a logic-based simulation method and a module for the power flow computation. The IEEE 30-bus system is used to evaluate the impact of attacks launched from outside or from within the substation networks. Countermeasures are identified for improvement of the cybersecurity.
A “Trap‐Release‐Amplify” Model of Chorus Waves Tao, Xin; Zonca, Fulvio; Chen, Liu
Journal of geophysical research. Space physics,
September 2021, Letnik:
126, Številka:
9
Journal Article
Recenzirano
Odprti dostop
Whistler mode chorus waves are quasi‐coherent electromagnetic emissions with frequency chirping. Various models have been proposed to understand the chirping mechanism, which is a long‐standing ...problem in space plasmas. Based on analysis of effective wave growth rate and electron phase space dynamics in a self‐consistent particle simulation, we propose a phenomenological model called the “Trap‐Release‐Amplify” (TaRA) model for chorus. In this model, phase space structures of correlated electrons are formed by nonlinear wave particle interactions, which mainly occur in the downstream of equator. When released from the wave packet in the upstream, these electrons lead to selective amplification of new emissions which satisfy the phase‐locking condition to maximize wave power transfer, resulting in frequency chirping. The phase‐locking condition at the release point gives a chirping rate that is, fully consistent with the one by Helliwell in case of a nonuniform background magnetic field. The nonlinear wave particle interaction part of the TaRA model results in a chirping rate that is, proportional to wave amplitude, a conclusion originally reached by Vomvoridis et al. Therefore, the TaRA model unifies two different results from seemingly unrelated studies. Furthermore, the TaRA model naturally explains fine structures of chorus waves, including subpackets and bandwidth, and their evolution through dynamics of phase‐trapped electrons. Finally, we suggest that this model could be applied to explain other related phenomena, including frequency chirping of chorus in a uniform background magnetic field and of electromagnetic ion cyclotron waves in the magnetosphere.
Key Points
We propose a new model to explain frequency chirping of chorus as selective amplification of new emissions by phase‐locking
The model unifies two different estimates of chorus chirping rate from seemingly unrelated studies
Fine structures of chorus and their dynamic evolution are explained by nonlinear electron phase space dynamics
PDF
