Abnormal Instances Research Articles

Enhancing SMOTE for imbalanced data with abnormal minority instances

Imbalanced datasets are frequent in machine learning, where certain classes are markedly underrepresented compared to others. This imbalance often results in sub-optimal model performance, as classifiers tend to favour the majority class. A significant challenge arises when abnormal instances, such as outliers, exist within the minority class, diminishing the effectiveness of traditional re-sampling methods like the Synthetic Minority Over-sampling Technique (SMOTE). This manuscript addresses this critical issue by introducing four SMOTE extensions: Distance ExtSMOTE, Dirichlet ExtSMOTE, FCRP SMOTE, and BGMM SMOTE. These methods leverage a weighted average of neighbouring instances to enhance the quality of synthetic samples and mitigate the impact of outliers. Comprehensive experiments conducted on diverse simulated and real-world imbalanced datasets demonstrate that the proposed methods improve classification performance compared to the original SMOTE and its most competitive variants. Notably, we demonstrate that Dirichlet ExtSMOTE outperforms most other proposed and existing SMOTE variants in terms of achieving better F1 score, MCC, and PR-AUC. Our results underscore the effectiveness of these advanced SMOTE extensions in tackling class imbalance, particularly in the presence of abnormal instances, offering robust solutions for real-world applications.

AnoOnly: Semi-supervised anomaly detection with the only loss on anomalies

Semi-supervised anomaly detection (SSAD) methods have demonstrated their effectiveness in enhancing unsupervised anomaly detection (UAD) by leveraging few-shot but instructive abnormal instances. However, the dominance of homogeneous normal data over anomalies biases the SSAD models against effectively perceiving anomalies. To address this issue and achieve balanced supervision between heavily imbalanced normal and abnormal data, we develop a novel framework called AnoOnly (Anomaly Only). Unlike existing SSAD methods that resort to strict loss supervision, AnoOnly suspends it for normal data and introduces a form of weak supervision. This weak supervision is instantiated through the utilization of batch normalization, which implicitly performs cluster learning on normal data. When integrated into existing SSAD methods, the proposed AnoOnly demonstrates remarkable performance enhancements across various models and datasets, achieving new state-of-the-art performance. Additionally, our AnoOnly is natively robust to label noise when suffering from data contamination. Our code is publicly available at https://github.com/cool-xuan/AnoOnly.

Advanced Anomaly Detection in ECG Signals Through Convolutional Autoencoders

This article aims to present a comprehensive study on convolutional autoencoders for advanced anomaly detection in ECG signals. Anomaly detection in complex datasets has become increasingly critical due to the rising need for systems that can effectively identify irregularities that may indicate fraud, system failures, or significant deviations from normal operations. Traditional methods often need help capturing nuanced patterns in high-dimensional data, necessitating more sophisticated approaches. This research uses an autoencoder-based model as a robust solution for anomaly detection, utilizing its capability to learn high-level representations in an unsupervised manner. The proposed model uses a convolutional autoencoder architecture to compress and decompress input data, thus highlighting anomalies through reconstruction errors. We outline detailed experiment strategies, including model training on average data to minimize reconstruction loss, setting an optimal threshold for anomaly sensitivity based on validation loss, and evaluating the model using precision, recall, F1-score, and AUC-ROC metrics. These experiments were conducted using a dataset with labeled normal and abnormal instances, allowing precise tuning and assessment of model performance. The results indicate that the autoencoder discriminates between normal and abnormal data, achieving high precision and recall at 99.22% and 98.98%, respectively. The confusion matrix and loss distribution analysis further validate the model's efficacy, clearly distinguishing between normal and abnormal data loss values concerning the defined threshold. This research shows the autoencoder model demonstrates high accuracy in anomaly detection and offers insights into the types of anomalies it can detect, supporting its application across various domains requiring reliable anomaly identification.

Utilizing IoMT-Based Smart Gloves for Continuous Vital Sign Monitoring to Safeguard Athlete Health and Optimize Training Protocols.

This paper presents the development of a vital sign monitoring system designed specifically for professional athletes, with a focus on runners. The system aims to enhance athletic performance and mitigate health risks associated with intense training regimens. It comprises a wearable glove that monitors key physiological parameters such as heart rate, blood oxygen saturation (SpO2), body temperature, and gyroscope data used to calculate linear speed, among other relevant metrics. Additionally, environmental variables, including ambient temperature, are tracked. To ensure accuracy, the system incorporates an onboard filtering algorithm to minimize false positives, allowing for timely intervention during instances of physiological abnormalities. The study demonstrates the system's potential to optimize performance and protect athlete well-being by facilitating real-time adjustments to training intensity and duration. The experimental results show that the system adheres to the classical "220-age" formula for calculating maximum heart rate, responds promptly to predefined thresholds, and outperforms a moving average filter in noise reduction, with the Gaussian filter delivering superior performance.

Rapid detection technology of sewing thread based on DeblurGAN-BSV3 defuzzification algorithm and ST-FPN detection algorithm

Stitch defect detection is an important means to ensure the quality of stitched fabrics. In order to realize the intellectualization of the sewing process, a rapid defect detection method based on machine vision was proposed. In order to solve the problem of motion blur caused by sewing speed, an image DeblurGAN-BSV3 algorithm is proposed. Firstly, different attention fusion mechanisms are introduced to achieve accurate extraction of image features. Secondly, the depth-separable convolution of the trunk network is replaced by blueprint convolution to ensure the processing speed of the detection algorithm and the clarity of the output image. In order to solve the difficulty of rapid detection of pin defects, a student–teacher feature pyramid matching method was proposed. Firstly, the abnormal region interpretation technology was used to explicitly visualize the defect area of the identified abnormal instances, and then the abnormal detection score was performed on the detected images to make a quality judgment. The experimental results show that the combination of the above two methods has a good ability for rapid defect detection. Rapid detection of common pin defects is realized.

Enhancing post-traumatic stress disorder patient assessment: leveraging natural language processing for research of domain criteria identification using electronic medical records

BackgroundExtracting research of domain criteria (RDoC) from high-risk populations like those with post-traumatic stress disorder (PTSD) is crucial for positive mental health improvements and policy enhancements. The intricacies of collecting, integrating, and effectively leveraging clinical notes for this purpose introduce complexities.MethodsIn our study, we created a natural language processing (NLP) workflow to analyze electronic medical record (EMR) data and identify and extract research of domain criteria using a pre-trained transformer-based natural language model, all-mpnet-base-v2. We subsequently built dictionaries from 100,000 clinical notes and analyzed 5.67 million clinical notes from 38,807 PTSD patients from the University of Pittsburgh Medical Center. Subsequently, we showcased the significance of our approach by extracting and visualizing RDoC information in two use cases: (i) across multiple patient populations and (ii) throughout various disease trajectories.ResultsThe sentence transformer model demonstrated high F1 macro scores across all RDoC domains, achieving the highest performance with a cosine similarity threshold value of 0.3. This ensured an F1 score of at least 80% across all RDoC domains. The study revealed consistent reductions in all six RDoC domains among PTSD patients after psychotherapy. We found that 60.6% of PTSD women have at least one abnormal instance of the six RDoC domains as compared to PTSD men (51.3%), with 45.1% of PTSD women with higher levels of sensorimotor disturbances compared to men (41.3%). We also found that 57.3% of PTSD patients have at least one abnormal instance of the six RDoC domains based on our records. Also, veterans had the higher abnormalities of negative and positive valence systems (60% and 51.9% of veterans respectively) compared to non-veterans (59.1% and 49.2% respectively). The domains following first diagnoses of PTSD were associated with heightened cue reactivity to trauma, suicide, alcohol, and substance consumption.ConclusionsThe findings provide initial insights into RDoC functioning in different populations and disease trajectories. Natural language processing proves valuable for capturing real-time, context dependent RDoC instances from extensive clinical notes.

CoMadOut—a robust outlier detection algorithm based on CoMAD

Unsupervised learning methods are well established in the area of anomaly detection and achieve state of the art performances on outlier datasets. Outliers play a significant role, since they bear the potential to distort the predictions of a machine learning algorithm on a given dataset. Especially among PCA-based methods, outliers have an additional destructive potential regarding the result: they may not only distort the orientation and translation of the principal components, they also make it more complicated to detect outliers. To address this problem, we propose the robust outlier detection algorithm CoMadOut, which satisfies two required properties: (1) being robust towards outliers and (2) detecting them. Our CoMadOut outlier detection variants using comedian PCA define, dependent on its variant, an inlier region with a robust noise margin by measures of in-distribution (variant CMO) and optimized scores by measures of out-of-distribution (variants CMO*), e.g. kurtosis-weighting by CMO+k. These measures allow distribution based outlier scoring for each principal component, and thus, an appropriate alignment of the degree of outlierness between normal and abnormal instances. Experiments comparing CoMadOut with traditional, deep and other comparable robust outlier detection methods showed that the performance of the introduced CoMadOut approach is competitive to well established methods related to average precision (AP), area under the precision recall curve (AUPRC) and area under the receiver operating characteristic (AUROC) curve. In summary our approach can be seen as a robust alternative for outlier detection tasks.

A New Particle Swarm Optimization Algorithm for Outlier Detection: Industrial Data Clustering in Wire Arc Additive Manufacturing

In this paper, a novel outlier detection method is proposed for industrial data analysis based on the fuzzy C-means (FCM) algorithm. An adaptive switching randomly perturbed particle swarm optimization algorithm (ASRPPSO) is put forward to optimize the initial cluster centroids of the FCM algorithm. The superiority of the proposed ASRPPSO is demonstrated over five existing PSO algorithms on a series of benchmark functions. To illustrate its application potential, the proposed ASRPPSO-based FCM algorithm is exploited in the outlier detection problem for analyzing the real-world industrial data collected from a wire arc additive manufacturing pilot line in Sweden. Experimental results demonstrate that the proposed ASRPPSO-based FCM algorithm outperforms the standard FCM algorithm in detecting outliers of real-world industrial data. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Electric arc (which is governed by the current and arc voltage) plays a significant role in monitoring the operating status of the wire arc additive manufacturing (WAAM) process. The nominal periodic current and voltage may occasionally change abruptly due to anomalies (such as arc instability, unstable metal transfer, geometrical deviations, and surface contaminations), which would affect the quality of the fabricated component. This paper focuses on detecting possible anomalies by analyzing the current and voltage during the WAAM process. A novel clustering-based outlier detection method is proposed for anomaly detection where abnormal and normal instances are categorized into two separate clusters. A new particle swarm optimization algorithm is put forward to optimize the initial cluster centroid so as to improve the detection accuracy. The proposed outlier detection method is applied to real-world data collected from a WAAM pilot line for detecting abnormal instances. Experimental results demonstrate the effectiveness of the proposed outlier detection method. The proposed outlier detection method can be applied to other industrial applications including electrical engineering, mechanical engineering and medical engineering. In the future, we aim to develop an online outlier detection system based on the proposed method for real-time for anomaly detection and defect prediction.

Attention-based early warning framework for abnormal operating conditions in fluid catalytic cracking units

Fluid Catalytic Cracking Unit (FCCU) is a critical processing technology in the oil refining industry, playing a vital role in energy efficiency and environmental protection. However, FCCU often encounters various abnormal operating conditions, leading to safety hazards, downtime, and reduced production efficiency. Early warning of these abnormal conditions is crucial but challenging due to high noise, strong hysteresis, and class imbalance problems. To tackle these challenges, a novel and universal attention-based framework called AEW-AOC (Attention-based Early Warning for Abnormal Operating Conditions) is specifically designed for FCCU applications. The proposed AEW-AOC framework incorporates three key components: (1) a Self-Correlation Denoiser (SCD) module is proposed to exploit spatiotemporal data correlation to effectively reduce noise; (2) a Convolutional Long Short-Term Memory (Conv-LSTM) module is employed to address the issue of strong hysteresis by capturing temporal variation features of process parameters; (3) an Anomaly Pattern Attention (APA) module is proposed to enhance the distinguishability of abnormal operating conditions based on clustering results from historical abnormal instances. Extensive experiments demonstrate the effectiveness and superiority of the proposed AEW-AOC framework, particularly in practical applications. Specifically, the AEW-AOC framework obtains an impressive fβ score of 91.00% on LIC201, 90.45% on LIC202, and 90.64% on LIC801. The proposed AEW-AOC framework shows great potential in enhancing safety, reducing downtime, optimizing efficiency, promoting sustainability, and expanding its applicability beyond FCCU. Its proactive and versatile nature makes it a valuable tool for improving industrial processes and driving advancements in the field of abnormal operating condition detection and prevention.

Investigating Outlier Detection Techniques Based on Kernel Rough Clustering

Background: Data quality is crucial to the success of big data analytics. However, the presence of outliers affects data quality and data analysis. Employing effective outlier detection techniques to eliminate dirty data can improve data quality and garner more accurate analytical insights. Data uncertainty presents a significant challenge for outlier detection methods and warrants further refinement in the era of big data. Objective: The unsupervised outlier detection based on the integration of clustering and outlier scoring scheme is the current research hotspot. However, hard clustering fails when dealing with abnormal patterns with uncertain and unexpected behavior. Rough boundaries help identify more accurate cluster structures. Therefore, this article uses uncertainty soft clustering based on rough set theory to extend the clustering technology and designs appropriate scoring schemes to capture abnormal instances. This solves the problem of outlier detection in uncertain and nonlinear complex data. Methods: This paper proposes the flow of an outlier detection algorithm based on Kernel Rough Clustering and then compares the detection accuracy with five existing popular methods using synthetic and real-world datasets. The results show that the proposed method has higher detection accuracy. Results: The detection precision and recall of the proposed method were improved. For the detection accuracy, it is superior to popular methods, indicating that the proposed method has a good detection effect in identifying outlier. Conclusion: Compared with popular methods, the proposed method has a slight advantage in detection accuracy and is one of the effective algorithms that can be selected for outlier detection.

Cognitive Refined Augmentation for Video Anomaly Detection in Weak Supervision.

Weakly supervised video anomaly detection is a methodology that assesses anomaly levels in individual frames based on labeled video data. Anomaly scores are computed by evaluating the deviation of distances derived from frames in an unbiased state. Weakly supervised video anomaly detection encounters the formidable challenge of false alarms, stemming from various sources, with a major contributor being the inadequate reflection of frame labels during the learning process. Multiple instance learning has been a pivotal solution to this issue in previous studies, necessitating the identification of discernible features between abnormal and normal segments. Simultaneously, it is imperative to identify shared biases within the feature space and cultivate a representative model. In this study, we introduce a novel multiple instance learning framework anchored on a memory unit, which augments features based on memory and effectively bridges the gap between normal and abnormal instances. This augmentation is facilitated through the integration of an multi-head attention feature augmentation module and loss function with a KL divergence and a Gaussian distribution estimation-based approach. The method identifies distinguishable features and secures the inter-instance distance, thus fortifying the distance metrics between abnormal and normal instances approximated by distribution. The contribution of this research involves proposing a novel framework based on MIL for performing WSVAD and presenting an efficient integration strategy during the augmentation process. Extensive experiments were conducted on benchmark datasets XD-Violence and UCF-Crime to substantiate the effectiveness of the proposed model.

Anomaly Detection Based on a 3D Convolutional Neural Network Combining Convolutional Block Attention Module Using Merged Frames.

With the recent rise in violent crime, the real-time situation analysis capabilities of the prevalent closed-circuit television have been employed for the deterrence and resolution of criminal activities. Anomaly detection can identify abnormal instances such as violence within the patterns of a specified dataset; however, it faces challenges in that the dataset for abnormal situations is smaller than that for normal situations. Herein, using datasets such as UBI-Fights, RWF-2000, and UCSD Ped1 and Ped2, anomaly detection was approached as a binary classification problem. Frames extracted from each video with annotation were reconstructed into a limited number of images of 3×3, 4×3, 4×4, 5×3 sizes using the method proposed in this paper, forming an input data structure similar to a light field and patch of vision transformer. The model was constructed by applying a convolutional block attention module that included channel and spatial attention modules to a residual neural network with depths of 10, 18, 34, and 50 in the form of a three-dimensional convolution. The proposed model performed better than existing models in detecting abnormal behavior such as violent acts in videos. For instance, with the undersampled UBI-Fights dataset, our network achieved an accuracy of 0.9933, a loss value of 0.0010, an area under the curve of 0.9973, and an equal error rate of 0.0027. These results may contribute significantly to solve real-world issues such as the detection of violent behavior in artificial intelligence systems using computer vision and real-time video monitoring.

End-to-End Transferable Anomaly Detection via Multi-Spectral Cross-Domain Representation Alignment

Anomaly detection (AD) aims to distinguish abnormal instances from what is defined as normal, which strongly correlates with the safe and robust applications of machine learning. A well-performed anomaly detector often relies on the training on massive labeled data, while it is of high cost to annotate data in practice. Fortunately, this dilemma can be solved by transferring the knowledge of a label-rich dataset (source domain) to assist the learning on the label-scarce dataset (target domain), which is known as domain adaptation in transfer learning. In this paper, we propose a Multi-spectral Cross-domain Representation Alignment (MsRA) method for the anomaly detection in the domain adaptation setting, where we can only access normal source data and limited normal target data. Specifically, MsRA first constructs multi-spectral feature representations by fusing different frequency components of the original features, which mitigates the information scarcity due to limited target training data by capturing richer input pattern information. Then we employ the adversarial training strategy to learn domain-invariant features and force the features of normal data to be more compact by the center clustering. Finally, the distance of each sample to the prototype of normal class can be used as its anomaly score, where the prototype is the center of both source and target data. In this way, we achieve anomaly detection in an end-to-end manner, without two-stage training for feature extraction and anomaly detection. Comprehensive experiments on cross-domain anomaly detection benchmarks validate the effectiveness of MsRA.

Abnormal Detection Using Two-Stage Method in Combined Power Plant

Complex systems, such as power plants, demand precise and reliable anomaly detection mechanisms. Traditional supervised learning approaches often fall short due to the challenges of imbalanced data and the scarcity of labeled abnormal instances. This paper introduces a two-stage methodology to address these challenges. The first stage emphasizes feature engineering, mitigating redundant sensor effects, and reducing dimensionality through Kmeans clustering and PCA. The second stage employs an LSTM-Autoencoder for abnormal event detection. Validated using data from a combined power plant, our approach demonstrates superior performance over existing techniques in terms of accuracy and computational efficiency. This research not only advances the field of anomaly detection in power plants but also offers insights for other complex systems.

Pregnancy Outcomes of Assisted Reproductive Technology (ART) Cycle Complicated by Ovarian Hyperstimulation Syndrome (OHSS): Case Series Study.

Ovarian hyperstimulation syndrome (OHSS) is a frequent, potentially lethal side effect of assisted reproductive technology (ART), distinguished by symptoms such as ovarian enlargement, ascites, and pleural effusion. This study is designed to study the effect of assisted reproductive technology (ART) cycle complicated by OHSS on pregnancy outcomes. A case series study at King Abdulaziz Medical City (KAMC) in Riyadh, Saudi Arabia, was executed to examine the pregnancy outcomes in in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) cycles. Fifteen patients were admitted to the IVF unit between January 2015 and December 2021. Data were retrieved from patients' medical records, and descriptive statistical methods were employed to analyze participants' data. The study assessed pregnancy outcomes for 15 female participants (mean age=31.1 years, SD=3.46) with a BMI range of 20-40 (mean BMI=29.6, SD=6.4), of whom 33.3% were classified as obese. The primary factor of infertility was anovulation (66.7%), followed by male factors (20%). About 26.7% of those affected by OHSS had moderate OHSS, and 73.3% had severe OHSS, with 100% of those with severe OHSS having undergone three embryo transfers. None of the participants developed gestational diabetes mellitus (DM), but one participant had high blood sugar levels (6.67% of total participants), with a mean glucose of 6.3±2.0. There were no instances of preeclampsia, gestational hypertension, abnormal placentas, or congenital abnormalities in newborns among the participants. Preterm deliveries were common, with 33.3% delivering between 32 and 37 weeks, 6.7% before 28 weeks, and 33.3% within 28-32 weeks. Overall, 73.3% of the participants experienced pregnancy, and the birth mode was almost evenly split between vaginal and cesarean birth. In conclusion, this research provides an exploration into the outcomes of pregnancies in women undergoing assisted reproductive technology treatments complicated by ovarian hyperstimulation syndrome. It shows anovulation as a prevalent cause of infertility and a noteworthy incidence of severe OHSS. Despite these challenges, a significant number of women were able to experience pregnancy, although preterm deliveries and abortions were common. The delivery methods were fairly balanced between vaginal birth and cesarean section. These findings underscore the necessity for more effective strategies to manage OHSS and improve pregnancy outcomes in ART procedures.

Detecting Multivariate Time Series Anomalies with Zero Known Label

Multivariate time series anomaly detection has been extensively studied under the one-class classification setting, where a training dataset with all normal instances is required. However, preparing such a dataset is very laborious since each single data instance should be fully guaranteed to be normal. It is, therefore, desired to explore multivariate time series anomaly detection methods based on the dataset without any label knowledge. In this paper, we propose MTGFlow, an unsupervised anomaly detection approach forMultivariate Time series anomaly detection via dynamic Graph and entityaware normalizing Flow, leaning only on a widely accepted hypothesis that abnormal instances exhibit sparse densities than the normal. However, the complex interdependencies among entities and the diverse inherent characteristics of each entity pose significant challenges to density estimation, let alone to detect anomalies based on the estimated possibility distribution. To tackle these problems, we propose to learn the mutual and dynamic relations among entities via a graph structure learning model, which helps to model the accurate distribution of multivariate time series. Moreover, taking account of distinct characteristics of the individual entities, an entity-aware normalizing flow is developed to describe each entity into a parameterized normal distribution, thereby producing fine-grained density estimation. Incorporating these two strategies, MTGFlow achieves superior anomaly detection performance. Experiments on five public datasets with seven baselines are conducted, MTGFlow outperforms the SOTA methods by up to 5.0 AUROC%.

An Optimal Modified Faster Region CNN Model for Diagnosis of Liver Diseases from Ultrasound Images

In recent years, the rate of mortality and morbidity increased due to uncontrolled conditions of liver diseases. Liver disorders are mainly caused because of inhalation of toxic gases, consumption of contaminated food, alcohol usage, etc. If the progression of liver disease is not detected at an early stage, it seriously threatens human life. Early diagnosis and timely treatment are the only ways to prevent such life-threatening diseases. This paper mainly aims to design an automated prediction model using ultrasound liver images. Due to non-invasive, cost-effective, and real-time imaging abilities, the ultrasound becomes most preferable modality. Therefore, the main purpose of the paper is at predicting liver diseases using the proposed modified faster region-based convolutional neural network-based hybrid grasshopper tunicate swarm (MFRCNN-HGTS) approach. The distortions in the raw ultrasound liver images are removed and the number of training images is increased by the augmentation process. The augmented data samples are then allowed to pass through a modified faster region-based convolutional neural network (MFRCNN) classifier to predict normal and abnormal instances from the images. To improve classification accuracy, the weight factors are assigned with appropriate values using the hybrid grasshopper tunicate swarm (HGTS) algorithm. The proposed MFRCNN-HGTS approach predicts liver diseases accurately and classifies them based on texture features as normal, cirrhosis, or hepatitis. The experimental setup is performed using MATLAB R2016a software. The proposed MFRCNN-HGTS method is evaluated by comparing it with other traditional methods. The analytic result illustrates that the proposed MFRCNN-HGTS techniques achieve 98.45% accuracy over other techniques.

Taste the rainbow: A review of color abnormalities affecting the herpetofauna of the British Isles

Over the years, the terminology in regards to the abnormal coloration of reptiles and amphibians has become more complex with not all authors in agreement regarding the different terms. This, combined with the diversity of chromatic abnormalities, has led to some confusion, particularly between hobbyists and conservationists who tend to use different technical jargon. In this review, we aim to tackle this issue by explaining how color within the skin of amphibians and reptiles arises, and evaluating which terminology should be used. This information will then be used to explore each of the known chromatic abnormalities observed in amphibians and reptiles before summarizing the known cases throughout the British Isles. Finally, we also present a number of previously unrecorded instances of color abnormalities in the hope that it promotes further examples to be recorded.

Anomaly detection using spatial and temporal information in multivariate time series

Real-world industrial systems contain a large number of interconnected sensors that generate a significant amount of time series data during system operation. Performing anomaly detection on these multivariate time series data can timely find faults, prevent malicious attacks, and ensure these systems safe and reliable operation. However, the rarity of abnormal instances leads to a lack of labeled data, so the supervised machine learning methods are not applicable. Furthermore, most current techniques do not take full advantage of the spatial and temporal dependencies implied among multiple variables to detect anomalies. Hence, we propose STADN, a novel Anomaly Detection Network Using Spatial and Temporal Information. STADN models the relationship graph between variables for a graph attention network to capture the spatial dependency between variables and utilizes a long short-term memory network to mine the temporal dependency of time series to fully use the spatial and temporal information of multivariate time series. STADN predicts the future behavior of each sensor by combining the historical behavior of the sensor and its neighbors, then detects and locates anomalies according to the prediction error. Furthermore, we improve the proposed model’s ability to discriminate anomaly and regularity and expand the prediction error gap between normal and abnormal instances by reconstructing the prediction errors. We conduct experiments on two real-world datasets, and the experimental results suggested that STADN achieves state-of-the-art outperformance.

Transferable visual pattern memory network for domain adaptation in anomaly detection

Anomaly detection transfer aims to utilize knowledge learned from source anomaly detection task to improve the performance of target anomaly detection task. Conventional methods typically assume that labeled normal or abnormal data are available in the source or target domain. However, many real-world applications do not satisfy this assumption because such labels are hard to collect. This study focuses on the case where anomalous labels are unavailable. More specifically, a rarely studied scenario in which the target domain contains unlabeled normal and abnormal instances, whereas only normal instances are available in the source domain, is addressed. To this end, a transferable visual pattern memory network was designed to transfer knowledge for anomaly detection tasks. The network comprises an adversarial domain adaptation method to extract transferable visual patterns, and a memory module utilized to store these patterns. The model utilizes transferable patterns stored in memory to identify anomalous samples. Moreover, a self-supervised objective is integrated to enhance the discriminability of target abnormal instances, thereby improving the anomaly detection performance. The results of extensive experiments conducted on publicly available anomaly-detection datasets verified the efficacy of the proposed approach.