Time series anomaly detection is a task of significant importance and has been widely employed in realistic scenarios. Most of existing methods conduct time series anomaly detection in an ...unsupervised manner, ignoring the limited number of labeled anomalies that are commonly available in practical situations. However, how to take advantage of these limited but valuable labeled anomalies to benefit time series anomaly detection requires fully exploration. To this end, we propose a weakly supervised time series anomaly detection method with dual-association discrepancy to effectively identify anomalies. Specifically, the proposed method utilizes the limited number of anomalies to enlarge the distance between the association discrepancy of the anomaly and that of the normal one, enforcing significant differences between abnormal and normal samples in discriminant places. We also design a masking strategy to enrich the representations of anomalies in latent feature space. Experiments with public available datasets have demonstrated the effectiveness of the proposed method.
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 recent years, sensor networks have been widely used for data and information collection. In order to alleviate or solve the energy problem of sensor networks, some researchers have proposed to use ...mobile devices to carry electrical energy, and using wireless energy transmission technology to transmit energy to sensor nodes, which led to the emergence of WRSNs. However, the deployment of mobile chargers greatly increases the construction cost of WRSNs. The focus of our research is how to reduce the total consumption of WRSN by reducing the number of mobile chargers. We achieve this goal by applying genetic algorithm (GA) to the path planning of mobile chargers.
Time series anomaly detection strives to uncover potential abnormal behaviors and patterns from temporal data, and has fundamental significance in diverse application scenarios. Constructing an ...effective detection model usually requires adequate training data stored in a centralized manner, however, this requirement sometimes could not be satisfied in realistic scenarios. As a prevailing approach to address the above problem, federated learning has demonstrated its power to cooperate with the distributed data available while protecting the privacy of data providers. However, it is still unclear that how existing time series anomaly detection algorithms perform with decentralized data storage and privacy protection through federated learning. To study this, we conduct a federated time series anomaly detection benchmark, named FedTADBench, which involves five representative time series anomaly detection algorithms and four popular federated learning methods. We would like to answer the following questions: (1)How is the performance of time series anomaly detection algorithms when meeting federated learning? (2) Which federated learning method is the most appropriate one for time series anomaly detection? (3) How do federated time series anomaly detection approaches perform on different partitions of data in clients? Numbers of results as well as corresponding analysis are provided from extensive experiments with various settings. The source code of our benchmark is publicly available at https://github.com/fanxingliu2020/FedTADBench.
FedTADBench: Federated Time-series Anomaly Detection Benchmark Liu, Fanxing; Zeng, Cheng; Zhang, Le ...
2022 IEEE 24th Int Conf on High Performance Computing & Communications; 8th Int Conf on Data Science & Systems; 20th Int Conf on Smart City; 8th Int Conf on Dependability in Sensor, Cloud & Big Data Systems & Application (HPCC/DSS/SmartCity/DependSys),
2022-Dec.
Conference Proceeding
Time series anomaly detection strives to uncover potential abnormal behaviors and patterns from temporal data, and has fundamental significance in diverse application scenarios. Constructing an ...effective detection model usually requires adequate training data stored in a centralized manner, however, this requirement sometimes could not be satisfied in realistic scenarios. As a prevailing approach to address the above problem, federated learning has demonstrated its power to cooperate with the distributed data available while protecting the privacy of data providers. However, it is still unclear that how existing time series anomaly detection algorithms perform with decentralized data storage and privacy protection through federated learning. To study this, we conduct a federated time series anomaly detection benchmark, named FedTADBench, which involves five representative time series anomaly detection algorithms and four popular federated learning methods. We would like to answer the following questions: (l)How is the performance of time series anomaly detection algorithms when meeting federated learning? (2) Which federated learning method is the most appropriate one for time series anomaly detection? (3) How do federated time series anomaly detection approaches perform on different partitions of data in clients? Numbers of results as well as corresponding analysis are provided from extensive experiments with various settings. The source code of our benchmark is publicly available at https://github.com/fanxindiu2020/FedTADBench.
Effective imputation is a crucial preprocessing step for time series analysis. Despite the development of numerous deep learning algorithms for time series imputation, the community lacks ...standardized and comprehensive benchmark platforms to effectively evaluate imputation performance across different settings. Moreover, although many deep learning forecasting algorithms have demonstrated excellent performance, whether their modeling achievements can be transferred to time series imputation tasks remains unexplored. To bridge these gaps, we develop TSI-Bench, the first (to our knowledge) comprehensive benchmark suite for time series imputation utilizing deep learning techniques. The TSI-Bench pipeline standardizes experimental settings to enable fair evaluation of imputation algorithms and identification of meaningful insights into the influence of domain-appropriate missingness ratios and patterns on model performance. Furthermore, TSI-Bench innovatively provides a systematic paradigm to tailor time series forecasting algorithms for imputation purposes. Our extensive study across 34,804 experiments, 28 algorithms, and 8 datasets with diverse missingness scenarios demonstrates TSI-Bench's effectiveness in diverse downstream tasks and potential to unlock future directions in time series imputation research and analysis. The source code and experiment logs are available at https://github.com/WenjieDu/AwesomeImputation.
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.
The potential applications of covalent organic frameworks (COFs) can be further developed by encapsulating functional nanoparticles within the frameworks. However, the synthesis of monodispersed ...core@shell structured COF nanocomposites without agglomeration remains a significant challenge. Herein, we present a versatile dual-ligand assistant strategy for interfacial growth of COFs on the functional nanoparticles with abundant physicochemical properties. Regardless of the composition, geometry or surface properties of the core, the obtained core@shell structured nanocomposites with controllable shell-thickness are very uniform without agglomeration. The derived bowl-shape, yolk@shell, core@satellites@shell nanostructures can also be fabricated delicately. As a promising type of photosensitizer for photodynamic therapy (PDT), the porphyrin-based COFs were grown onto upconversion nanoparticles (UCNPs). With the assistance of the near-infrared (NIR) to visible optical property of UCNPs core and the intrinsic porosity of COF shell, the core@shell nanocomposites can be applied as a nanoplatform for NIR-activated PDT with deep tissue penetration and chemotherapeutic drug delivery.
The synthesis of crystalline two-dimensional polymers (2DPs) with proper bandgaps and well-defined repeating units presents a great challenge to synthetic chemists. Here we report the first solution ...synthesis of a single-layer/few-layer triazine-based 2DP via trimerization of carbonitrile at the interface of dichloromethane and trifluoromethanesulfonic acid. The processable triazine-based 2DP can be assembled into mechanically strong layered free-standing films with a high specific surface area via filtration. Moreover, the highly crystalline triazine-based 2DP can function as the active semiconducting layer in a field-effect transistor via drop coating and exhibits slightly bipolar behavior with a high on/off ratio of 103 and a remarkable mobility of 0.15 cm2 V–1 s–1.
Abstract
Metallic tungsten disulfide (WS
2
) monolayers have been demonstrated as promising electrocatalysts for hydrogen evolution reaction (HER) induced by the high intrinsic conductivity, however, ...the key challenges to maximize the catalytic activity are achieving the metallic WS
2
with high concentration and increasing the density of the active sites. In this work, single-atom-V catalysts (V SACs) substitutions in 1T-WS
2
monolayers (91% phase purity) are fabricated to significantly enhance the HER performance via a one-step chemical vapor deposition strategy. Atomic-resolution scanning transmission electron microscopy (STEM) imaging together with Raman spectroscopy confirm the atomic dispersion of V species on the 1T-WS
2
monolayers instead of energetically favorable 2H-WS
2
monolayers. The growth mechanism of V SACs@1T-WS
2
monolayers is experimentally and theoretically demonstrated. Density functional theory (DFT) calculations demonstrate that the activated V-atom sites play vital important role in enhancing the HER activity. In this work, it opens a novel path to directly synthesize atomically dispersed single-metal catalysts on metastable materials as efficient and robust electrocatalysts.