Weakly supervised anomaly detection aims at learning an anomaly detector from a limited amount of labeled data and abundant unlabeled data. Recent works build deep neural networks for anomaly ...detection by discriminatively mapping the normal samples and abnormal samples to different regions in the feature space or fitting different distributions. However, due to the limited number of annotated anomaly samples, directly training networks with the discriminative loss may not be sufficient. To overcome this issue, this article proposes a novel strategy to transform the input data into a more meaningful representation that could be used for anomaly detection. Specifically, we leverage an autoencoder to encode the input data and utilize three factors, hidden representation, reconstruction residual vector, and reconstruction error, as the new representation for the input data. This representation amounts to encode a test sample with its projection on the training data manifold, its direction to its projection, and its distance to its projection. In addition to this encoding, we also propose a novel network architecture to seamlessly incorporate those three factors. From our extensive experiments, the benefits of the proposed strategy are clearly demonstrated by its superior performance over the competitive methods. Code is available at: https://github.com/yj-zhou/Feature_Encoding_with_AutoEncoders_for_Weakly-supervised_Anomaly_Detection .
In this paper, the slide film damping (SFD) of micro-scale oscillatory Couette flow has been researched by an effective multiple relaxation time lattice Boltzmann method (MRT-LBM). The validity and ...effectiveness of MRT-LBM for solving SFD have been verified through contrasting the velocity distribution between the upper plate and the substrate of MRT-LBM and the direct simulation Monte Carlo (DSMC) model. The impacts of the vibration frequency and gap of plates on the damping are discussed, and the result shows that the nonlinear character of velocity profiles is manifest and SFD increases substantially for a larger vibration frequency in the oscillatory gas flow; SFD reduces obviously with the gap increases. Consequently, the results further confirm the implementation of LBM in analysis of the non-equilibrium micro-scale gas flow.
With the wide deployment of edge devices, a variety of emerging applications have been deployed at the edge of network. To guarantee the safe and efficient operations of the edge applications, ...especially the extensive web applications, it is important and challenging to detect packet payload anomalies, which can be expressed as a number of specific strings that may cause attacks. Although some approaches have achieved remarkable progress, they are with limited applications since these approaches are dependent on in-depth expert knowledge, e.g., signatures describing anomalies or communication protocol at the application level. Moreover, they might fail to detect the payload anomalies that may have long-term dependency relationships at the edge of network. To overcome these limitations and adaptively detect anomalies from packet payloads, we propose a deep learning based framework which does not rely on any in-depth expert knowledge and is capable of detecting anomalies that have long-term dependency relationships. The proposed framework consists of two parts. First, a novel block sequence construction method is proposed to obtain a valid expression of a payload. The block sequence could encapsulate both the high-dimension information and the underlying sequential information which facilitate the anomaly detection. Secondly, we design a detection model to learn two different dependency relationships within the block sequence, which is based on Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN) and Multi-head Self Attention Mechanism. Furthermore, we cast the anomaly detection as a classification problem and employ a classifier with attention mechanism to integrate information and detect anomalies. Extensive experimental results on three public datasets indicate that our model could achieve a higher detection rate, while keeping a lower false positive rate compared with two traditional machine learning methods and three state-of-the-art methods.
Weakly-supervised anomaly detection aims at learning an anomaly detector from a limited amount of labeled data and abundant unlabeled data. Recent works build deep neural networks for anomaly ...detection by discriminatively mapping the normal samples and abnormal samples to different regions in the feature space or fitting different distributions. However, due to the limited number of annotated anomaly samples, directly training networks with the discriminative loss may not be sufficient. To overcome this issue, this paper proposes a novel strategy to transform the input data into a more meaningful representation that could be used for anomaly detection. Specifically, we leverage an autoencoder to encode the input data and utilize three factors, hidden representation, reconstruction residual vector, and reconstruction error, as the new representation for the input data. This representation amounts to encode a test sample with its projection on the training data manifold, its direction to its projection and its distance to its projection. In addition to this encoding, we also propose a novel network architecture to seamlessly incorporate those three factors. From our extensive experiments, the benefits of the proposed strategy are clearly demonstrated by its superior performance over the competitive methods.
With the widespread adoption of cloud services, especially the extensive deployment of plenty of Web applications, it is important and challenging to detect anomalies from the packet payload. For ...example, the anomalies in the packet payload can be expressed as a number of specific strings which may cause attacks. Although some approaches have achieved remarkable progress, they are with limited applications since they are dependent on in-depth expert knowledge, e.g., signatures describing anomalies or communication protocol at the application level. Moreover, they might fail to detect the payload anomalies that have long-term dependency relationships. To overcome these limitations and adaptively detect anomalies from the packet payload, we propose a deep learning based framework which consists of two steps. First, a novel feature engineering method is proposed to obtain the block-based features via block sequence extraction and block embedding. The block-based features could encapsulate both the high-dimension information and the underlying sequential information which facilitate the anomaly detection. Second, a neural network is designed to learn the representation of packet payload based on Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNN). Furthermore, we cast the anomaly detection as a classification problem and stack a Multi-Layer Perception (MLP) on the above representation learning network to detect anomalies. Extensive experimental results on three public datasets indicate that our model could achieve a higher detection rate, while keeping a lower false positive rate compared with five state-of-the-art methods.
Achieving high energy density and long cycle life in realistic batteries is still an unmet need, which has triggered research into the discoveries of new electrode materials as well as new storage ...mechanisms. As a kind of new cathode materials for rechargeable lithium batteries, organosulfide compounds R‐Sn‐R (n = 3–6) based on conversion chemistries of SS bonds have many advantages and promising prospects; however, poor electric/ionic conductivity and sluggish redox kinetics is a major hinder for their applications. Here an organic–inorganic hybrid cathode by introducing 1T MoS2 grown on reduced graphene oxide to hybridize with phenyl tetrasulfide (Ph‐S4‐Ph, theoretical specific capacity 570 mAh g−1), enhancing the battery performance is reported. This includes the improved charge transfer, stable long cycles, and close‐to‐practical energy density in coin cells and pouch cells, which also show high mass loadings and contents, and low electrolyte dependence. Furthermore, the dynamic 1T‐2H mixed‐phase during the charge/discharge is revealed to be critical for the improved performance. This study proves the hybrid nanomaterials as a promising solution to address the challenges facing lithium‐organosulfide batteries.
An organic–inorganic hybrid cathode consisting of 1T phase MoS2 grown on reduced graphene oxide mixed with carbon nanotubes and phenyl tetrasulfide (PTS) retains 69.8% of the theoretical capacity of PTS after 950 cycles in rechargeable lithium batteries. The cells also show high mass loadings and low electrolyte dependence. The dynamic 1T‐2H mixed‐phases MoS2 is critical for the improved performance.
Achieving high energy density and long cycle life in realistic batteries is still an unmet need, which has triggered research into the discoveries of new electrode materials as well as new storage ...mechanisms. As a kind of new cathode materials for rechargeable lithium batteries, organosulfide compounds R-S
-R (n = 3-6) based on conversion chemistries of SS bonds have many advantages and promising prospects; however, poor electric/ionic conductivity and sluggish redox kinetics is a major hinder for their applications. Here an organic-inorganic hybrid cathode by introducing 1T MoS
grown on reduced graphene oxide to hybridize with phenyl tetrasulfide (Ph-S
-Ph, theoretical specific capacity 570 mAh g
), enhancing the battery performance is reported. This includes the improved charge transfer, stable long cycles, and close-to-practical energy density in coin cells and pouch cells, which also show high mass loadings and contents, and low electrolyte dependence. Furthermore, the dynamic 1T-2H mixed-phase during the charge/discharge is revealed to be critical for the improved performance. This study proves the hybrid nanomaterials as a promising solution to address the challenges facing lithium-organosulfide batteries.
Great efforts have been made to tackle the issues of the shuttle effect and kinetics hysteresis in lithium‐sulfur (Li−S) battery, but few on tuning the reaction path of sulfur cathode. Herein, we ...report a strategy to replace inorganic sulfur with liquid organosulfide and construct a novel liquid‐liquid interface between cathode and electrolyte, which effectively inhibits the shuttle effect and simplifies the solid‐liquid‐solid conversion reaction to only liquid‐solid process, thus greatly improving the reaction kinetics. The Li|PTS half‐cell exhibits excellent cycling stability at 0.5 C, with a capacity retention of 64.9 % after 750 cycles. The Li|PTS pouch cell with a high PTS loading of 3.1 g delivers a maximum capacity of 997 mAh and maintains 82.1 % of initial capacity after 50 cycles at the current of 100 mA. This work enriches the reaction mechanism of Li−S batteries and provides new insights for the development of interphase chemistry in the design of cathodes.
In view of the poor electronic/ionic conductivity of the solid‐liquid interface in Li−S battery, a strategy of replacing sulfur with liquid organosulfide (PTS) and constructing a novel liquid‐liquid interface between cathode and electrolyte using LHCE is reported, which effectively inhibits the shuttle effect and simplifies the reaction mechanism to only liquid‐solid process, thus obtaining a good reaction kinetics and cycling stability.
Drought is one of the most common abiotic stressors in plants. Melatonin (MT) is a high-efficiency and low-toxicity growth regulator that plays an important role in plant responses to drought stress. ...As a wild relative of wheat,
has become an important species for the improvement of degraded grasslands and the replanting of sandy grasslands. However, the physiological and molecular mechanisms by which exogenous MT regulates drought stress in
remain unclear. To assess the effectiveness of MT intervention (100 mg·L
), polyethylene glycol 6000 was used to simulate drought stress, and its ameliorating effects on drought stress in
seedlings were investigated through physiology, transcriptomics, and metabolomics. Physiological analysis indicated that MT treatment increased the relative water content and chlorophyll content and decreased the relative conductivity of
seedlings. Additionally, MT decreased malondialdehyde (MDA) and reactive oxygen species (ROS) accumulation by enhancing antioxidant enzyme activities. The transcriptome and metabolite profiling analysis of
seedlings treated with and without MT under drought stress identified the presence of 13,466 differentially expressed genes (DEGs) and 271 differentially expressed metabolites (DEMs). The integrated analysis of transcriptomics and metabolomics showed that DEGs and DEMs participated in diverse biological processes, such as flavonoid biosynthesis and carbohydrate metabolism. Moreover, MT may be involved in regulating the correlation of DEGs and DEMs in flavonoid biosynthesis and carbohydrate metabolism during drought stress. In summary, this study revealed the physiological and molecular regulatory mechanisms of exogenous MT in alleviating drought stress in
seedlings, and it provides a reference for the development and utilization of MT and the genetic improvement of drought tolerance in plants from arid habitats.
The two-stroke engine is a common power source for small and medium-sized unmanned aerial vehicles (UAV), which has wide civil and military applications. To improve the engine performance, we chose a ...prototype two-stroke small areoengine, and optimized the geometric parameters of the scavenging ports by performing one-dimensional (1D) and three-dimensional (3D) computational fluid dynamics (CFD) coupling simulations. The prototype engine is tested on a dynamometer to measure in-cylinder pressure curves, as a reference for subsequent simulations. A GT Power simulation model is established and validated against experimental data to provide initial conditions and boundary conditions for the subsequent AVL FIRE simulations. Four parameters are considered as optimal design factors in this research: Tilt angle of the central scavenging port, tilt angle of lateral scavenging ports, slip angle of lateral scavenging ports, and width ratio of the central scavenging port. An evaluation objective function based on the Benson/Bradham model is selected as the optimization goal. Two different operating conditions, including the take-off and cruise of the UAV are considered. The results include: (1) Orthogonal experiments are analyzed, and the significance of parameters are discussed; (2) the best factors combination is concluded, followed by simulation verification; (3) results before and after optimization are compared in details, including specific scavenging indexes (delivery ratio, trapping efficiency, scavenging efficiency, etc.), conventional performance indicators, and the sectional views of gas composition distribution inside the cylinder.