Search

Detecting anomalies accurately in time series data has been receiving considerable attention due to its enormous potential for a wide array of applications. Numerous unsupervised anomaly detection methods for time series have been developed because of the difficulty of obtaining accurate labels. However, most existing unsupervised approaches suffer from the problem of anomaly contamination, which results in models that are unable to learn the normal pattern well and further deteriorate the performance of detection methods. To this end, a novel unsupervised method, called Self-adversarial Variational Autoencoder with Spectral Residual (SaVAE-SR), is introduced for time series anomaly detection in this paper. The SaVAE-SR first produces labels for unlabeled training data using the spectral residual technique to identify the most critical anomalies. A VAE model with a modified loss that can leverage label information to remove the influence of anomalous points is then trained in a self-adversarial manner, enabling the model to self-evaluate the learning of complex data distribution and improve itself accordingly. Specifically, the encoder acts as an encoder to approximate the posterior of latent variables and as a discriminator to evaluate the generative ability of the generator and improve itself accordingly. The generator is trained to capture the underlying data distribution and attempts to produce real samples to deceive the discriminator. The encoder and generator of the model compete with each other just like the behavior of GANs but work together under the theoretical framework of VAEs. As a result, the SaVAE-SR model combines the respective strengths of the VAE and adversarial training but does not require an additional discriminator, which makes the whole model very compact. Extensive experiments on five datasets demonstrate the superiority of the proposed method over the existing state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In intelligent transportation systems (ITS), deploying fine-grained sensors to continuously collect spatial–temporal traffic data is important but impractical due to the expensive cost. Fortunately, spatial–temporal kriging methods bring advanced solutions for interpolating traffic data for locations without sensors, but they still have the following two drawbacks: (1) The widely adopted predefined and adaptive graphs are either inflexible or limited to transductive learning. (2) The sampling strategies to support inductive learning on new graphs result in losing important partial information. To overcome the above issues, in this paper, we propose an Inductive and Adaptive Graph Convolution Networks (IAGCN) for spatial–temporal traffic data kriging in an inductive manner. Specifically, we propose an adaptive graph constructor to model the hidden spatial relation of nodes and learn the spatial dependency that the predefined graph structure cannot capture. It can work inductively and does not require retraining when introducing new nodes. Additionally, we design a framework that integrates inductive graph convolution and temporal convolution to simultaneously capture the complex spatial–temporal dependencies of traffic data. Finally, to address the information loss issue caused by random sampling, we design a predicted-based constraint task that perceives and utilizes the history information of all observed sensors to predict the current data, as well as approach the result of the interpolation and prediction. Experiments on four real-world datasets show that IAGCN outperforms the state-of-the-art baselines. [ABSTRACT FROM AUTHOR]

• Method achieved a good result of 94.4% accuracy for apnea detection in segment. • Firstly employ deep residual network for apnea detectionbased on low-sample-rate data. • Weights of network can be fine-tuned on clinical. • Strong adaptivity. Model also works well using ECG-derived respiration (EDR) signal. Apnea is one of the most common conditions that causes sleep-disorder breathing. With growing number of patients worldwide, more and more patients suffer from complications of apnea. But most of them stay untreated due to the complex and time-consuming polysomnography (PSG) diagnosis method. Effective and precise diagnosis support system using electrocardiograph (ECG) is required. In this paper, we propose an approach using residual network to detect apnea based on RR intervals (intervals between R-peaks of ECG signal). In our model, we apply residual network to represent information carried by RR intervals. Moreover, we proposed a novel perspective, called dynamic autoregressive representation, to provide interpretation of representing the RR intervals by convolutional layers. This approach is tested for per-segment apnea detection using publicly available dataset on Physionet. 30 overnight recordings are used for training and 5 for testing. We achieve a good result of 94.4% accuracy, 93.0% sensitivity and 94.9% specificity. This result outperform other prevalent methods based on RR intervals. This model also shows its good adaptivity while using ECG-derived respiration signal (EDR) in experiments. Its extensiveness is evaluated and compared in experiments. The proposed model is also compared with deep neural networks using original ECG signals for apnea detection, and it achieves better result using fewer input samples. We develop a deep residual network to detect apnea on low-sample-rate RR intervals. The result suggests a possibility of representing RR intervals by neural network. The model showed strong adaptivity when using EDR input. [ABSTRACT FROM AUTHOR]

Networks can represent a wide range of complex systems, such as social, biological and technological systems. Link prediction is one of the most important problems in network analysis, and has attracted much research interest recently. Many link prediction methods have been proposed to solve this problem with various techniques. We can note that clustering information plays an important role in solving the link prediction problem. In previous literatures, we find node clustering coefficient appears frequently in many link prediction methods. However, node clustering coefficient is limited to describe the role of a common-neighbor in different local networks, because it cannot distinguish different clustering abilities of a node to different node pairs. In this paper, we shift our focus from nodes to links, and propose the concept of asymmetric link clustering (ALC) coefficient. Further, we improve three node clustering based link prediction methods via the concept of ALC. The experimental results demonstrate that ALC-based methods outperform node clustering based methods, especially achieving remarkable improvements on food web, hamster friendship and Internet networks. Besides, comparing with other methods, the performance of ALC-based methods are very stable in both globalized and personalized top-L link prediction tasks. [ABSTRACT FROM AUTHOR]

Multiscale entropy analysis has become a prevalent complexity measurement and been successfully applied in various fields. However, it only takes into account the information of mean values (first moment) in coarse-graining procedure. Then generalized multiscale entropy (MSE n ) considering higher moments to coarse-grain a time series was proposed and MSE σ 2 has been implemented. However, the MSE σ 2 sometimes may yield an imprecise estimation of entropy or undefined entropy, and reduce statistical reliability of sample entropy estimation as scale factor increases. For this purpose, we developed the refined model, RMSE σ 2 , to improve MSE σ 2 . Simulations on both white noise and 1 ∕ f noise show that RMSE σ 2 provides higher entropy reliability and reduces the occurrence of undefined entropy, especially suitable for short time series. Besides, we discuss the effect on RMSE σ 2 analysis from outliers, data loss and other concepts in signal processing. We apply the proposed model to evaluate the complexity of heartbeat interval time series derived from healthy young and elderly subjects, patients with congestive heart failure and patients with atrial fibrillation respectively, compared to several popular complexity metrics. The results demonstrate that RMSE σ 2 measured complexity (a) decreases with aging and diseases, and (b) gives significant discrimination between different physiological/pathological states, which may facilitate clinical application. [ABSTRACT FROM AUTHOR]

Predicting missing links in incomplete complex networks efficiently and accurately is still a challenging problem. The recently proposed Cannistrai–Alanis–Ravai (CAR) index shows the power of local link/triangle information in improving link-prediction accuracy. Inspired by the idea of employing local link/triangle information, we propose a new similarity index with more local structure information. In our method, local link/triangle structure information can be conveyed by clustering coefficient of common-neighbors directly. The reason why clustering coefficient has good effectiveness in estimating the contribution of a common-neighbor is that it employs links existing between neighbors of a common-neighbor and these links have the same structural position with the candidate link to this common-neighbor. In our experiments, three estimators: precision, AUP and AUC are used to evaluate the accuracy of link prediction algorithms. Experimental results on ten tested networks drawn from various fields show that our new index is more effective in predicting missing links than CAR index, especially for networks with low correlation between number of common-neighbors and number of links between common-neighbors. [ABSTRACT FROM AUTHOR]

Abstract: The detection of overlapping community structure in networks can give insight into the structures and functions of many complex systems. In this paper, we propose a simple but efficient overlapping community detection method for very large real-world networks. Taking a high-quality, non-overlapping partition generated by existing, efficient, non-overlapping community detection methods as input, our method identifies overlapping nodes between each pair of connected non-overlapping communities in turn. Through our analysis on modularity, we deduce that, to become an overlapping node without demolishing modularity, nodes should satisfy a specific condition presented in this paper. The proposed algorithm outputs high quality overlapping communities by efficiently identifying overlapping nodes that satisfy the above condition. Experiments on synthetic and real-world networks show that in most cases our method is better than other algorithms either in the quality of results or the computational performance. In some cases, our method is the only one that can produce overlapping communities in the very large real-world networks used in the experiments. [Copyright &y& Elsevier]

Complex physiological signals carry critical information about the underlying dynamics of the complex system that produced them. By analyzing and exploring these signals, we can identify which state the system is in and understand the underlying dynamics of that system. Numerous linear and nonlinear methods have been developed for analyzing physiological signals, but they either rely on certain assumptions or are easily affected by various factors, which limits their applications. In this paper, we propose a novel dissimilarity measure between two signals, which is based on the ordinal pattern representation of signals. Therefore, it naturally inherits the advantages of ordinal pattern analysis, namely simplicity, robustness, and low complexity in computation without further prior assumptions. Finally, we apply this new measure to analyze several physiological signals derived from different physiological and pathological conditions to assess the effectiveness of the proposed measure. Experimental results from heart rate analysis, EEG signals analysis of healthy subjects and patients with epilepsy, and analysis of EEG signals at different sleep stages in healthy subjects demonstrate that the new measure is capable of distinguishing signals from different conditions and is therefore a useful tool for analyzing various physiological signals. • A new dissimilarity measure based on ordinal pattern representation is proposed. • The new measure is robust to noise and without further model assumptions. • The new measure is capable of distinguishing signals from different states. • Three case studies demonstrate the effectiveness of the new dissimilarity measure. [ABSTRACT FROM AUTHOR]

Distantly supervised relation extraction aims to obtain relational facts from unstructured texts. Although distant supervision can automatically generate labeled training instances, it inevitably suffers from the wrong-label problem. Most of the current work is based on the bag-level for solving the noise problem, where a bag is composed of multiple sentences containing mentions of the same entity pair. However, previous studies mostly represent sentences from a single perspective, wherein insufficient modeling of global information restricts the effectiveness of denoising. In this study, we propose a novel distantly supervised relation extraction approach that incorporates the global contextual information of sentences to guide the denoising process and generate an effective bag-level representation. Simultaneously, knowledge-aware word embeddings were generated to enrich sentence-level representations by introducing both structured knowledge from external knowledge graphs and semantic knowledge from the corpus. The experimental results demonstrate that our proposed approach outperforms state-of-the-art methods on both versions of the large-scale benchmark New York Times dataset. In addition, the differences between the two versions of the dataset were investigated through further comparative experiments. [ABSTRACT FROM AUTHOR]

Providing both point estimation and uncertainty quantification for traffic forecasting is crucial for supporting accurate and reliable services in intelligent transportation systems. However, the majority of existing traffic forecasting works mainly focus on point estimation without quantifying the uncertainty of predictions. Meanwhile, existing uncertainty quantification (UQ) methods fail to capture the inherent static characteristics of traffic uncertainty along both the spatial and temporal dimensions. Directly equipping the traffic forecasting works with uncertainty quantification techniques may even damage the prediction accuracy. In this paper, we propose a novel traffic forecasting model aiming at providing point estimation and uncertainty quantification simultaneously, called STUP. Compared to the traditional graph convolution networks (GCNs), our framework is able to incorporate uncertainty quantification into traffic forecasting to further improve forecasting performance. Specifically, we first develop an adaptive strategy to initialize uncertainty distribution. Then a kind of spatial–temporal uncertainty layer is carefully designed to model the evolution process of both the traffic state and its corresponding uncertainty, along with a gated adjusting unit to avoid error information propagation. Finally, we propose a novel constraint loss to further help improve the forecasting accuracy and to alleviate the training difficulty caused by the lack of uncertainty labels. Experiments on five real-world traffic datasets demonstrate that STUP outperforms the state-of-the-art baselines on both the traffic prediction task and uncertainty quantification task. [Display omitted] • A spatial–temporal traffic forecasting method considering uncertainty is proposed. • Discovering the relationship between traffic data and factors of uncertainty. • Balancing the accuracy and reliability on traffic forecasting. [ABSTRACT FROM AUTHOR]

Electrocardiogram (ECG) is a widely used diagnostic tool for arrhythmia assessment in clinical practice. However, current arrhythmia detection algorithms rely heavily on signal-based data, while cardiologists often use image-based data. This discrepancy, combined with individual differences in physiological signals, poses challenges for accurate arrhythmia detection. To address these challenges and improve arrhythmia detection performance, we propose a homologous and heterogeneous multi-view inter-patient adaptive network. We designed a multi-view representation learning module to capture dynamic and morphological characteristics from ECG signals and electrocardiographic images. Expert knowledge was also elicited to gain internally-invariant characteristics of each category. Finally, we designed a new loss function that aligns the embedding of the source and target domains in the feature space to minimize the negative effects of individual differences. Experiments on the MIT-BIH arrhythmia database demonstrate that our proposed method outperforms state-of-the-art baselines in terms of accuracy, positive predictive value, sensitivity and F1-score. These results indicate the effectiveness of our method in accurately detecting arrhythmias. Our homologous and heterogeneous multi-view inter-patient adaptive network successfully addresses the challenges of utilizing both ECG signal and electrocardiographic image data for arrhythmia detection and overcoming individual differences in physiological signals. Our proposed method has the potential to improve early diagnosis and treatment outcomes of arrhythmias in clinical practice. • A homologous and heterogeneous multi-view adaptive network is proposed. • The network captures dynamic and morphological characteristics of ECG. • The network elicits expert knowledge to model internally-invariant characteristics. • The network is able to overcome individual differences. • The proposed network outperforms the state-of-the-art baselines. [ABSTRACT FROM AUTHOR]

Traditional collaborative filtering based recommender systems generally suffer from the interaction data sparsity problem. Therefore, social recommendation is proposed to mitigate the issue and improve recommendation performance by introducing social information. Existing social recommendation studies primarily focus on the direct connections between users, such as friendship or users' correlation. Unfortunately, there often is a severe data sparsity issue in above social data as well, which limits the performance of these models. In contrast, user-group relationships, another valuable social information that is formed by users joining the groups they are interested in, have received insufficient attention. In this paper, we focus on this relationship, demonstrate its excellent effectiveness in alleviating the problem of data sparsity, and integrate it into our recommendation model IGRec (I ntegrating user- G roup relationships for social Rec ommendation) in a reasonable way. Specifically, to address the problem that existing group-information-enhanced methods have not modeled users' collaborative interests and social influence in depth, we reformulate the available data into two bipartite graphs: user-item graph and user-group graph. And then employ more robust high-order GCN-based model combining a multi-layer attention mechanism to learn user and item representation from two graphs. Furthermore, we notice that due to the high complexity of user-group networks, the interests of some users in the same group may be far different, especially in those large-scale groups. The indiscriminate use of high-order neighbors' information in user-group graph may result in the introduction of negative information during the embedding propagation. Thus, to obtain a more precise representation for user and item, we propose to constrain the graph convolution operations at the social side inside subgraphs composed of users with similar interests and the groups they have joined in our model. Finally, experimental results on three real-world datasets clearly show the effectiveness of our proposed model. [ABSTRACT FROM AUTHOR]

Correlated time series refer to multiple time series which are recorded simultaneously to monitor the changing of multiple observations in a whole system. Correlated time series prediction plays a significant role in many real-world applications to help people make reasonable decisions. Yet it is very challenging, because different from single time series, correlated time series show both intra-sequence temporal dependencies and inter-sequence spatial dependencies. In addition, correlated time series are also affected by external factors in actual scenarios. Although RNNs have been proved to be effective on sequential data modeling, existing related works only focus on sequential patterns in a single time series, failing to comprehensively consider the inter-dependencies among multiple time series, which is essential for correlated time series prediction. In this paper, we propose a novel variant of LSTM, named CTS-LSTM, to collectively forecast correlated time series. Specifically, spatial and temporal correlations are explicitly modeled and respectively maintained in cells to capture the complex non-linear patterns in correlated time series. A general interface for handling external factors is further designed to enhance forecasting performance of the model. Experiments are conducted on two types of real-world datasets, viz., civil aviation passenger demand data and air quality data. And our CTS-LSTM achieves at least 9.0%, 16.5% and 21.3% lower RMSE, MAE and MAPE compared to the state-of-the-art baselines. [ABSTRACT FROM AUTHOR]