Vector Autoregressive Model-Based Anomaly Detection in Aviation Systems

Melnyk, Igor; Matthews, Bryan; Valizadegan, Hamed; Banerjee, Arindam; Oza, Nikunj C.

doi:10.2514/1.i010394

Cited by 37 publications

(30 citation statements)

References 28 publications

(40 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Some works use (hidden) Markov chain models, e.g., to detect (groups of) sequences which show significant differences in terms of state transition probabilities [4,29,31,45]. Other methods use nearest-neighbours algorithms [32] or distance measures [47] to quantify how any given sequence s j differs from other instances in S. Adopting a collective definition of anomalies [16], a third class of methods is based on hypothesis testing techniques to detect outliers in the distribution of features of sequences [5,30,46]. Sequential pattern mining.…”

Section: Related Workmentioning

confidence: 99%

HYPA: Efficient Detection of Path Anomalies in Time Series Data on Networks

LaRock

Nanumyan

Scholtes

et al. 2020

Proceedings of the 2020 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

The unsupervised detection of anomalies in time series data has important applications, e.g., in user behavioural modelling, fraud detection, and cybersecurity. Anomaly detection has been extensively studied in categorical sequences. But we often have access to time series data that contain paths in networks. Examples include transaction sequences in financial networks, click streams of users in networks of cross-referenced documents, or travel itineraries in transportation networks. To reliably detect anomalies we must account for the fact that such data contain a large number of independent observations of short paths constrained by a graph topology. Moreover, the heterogeneity of real systems rules out frequencybased anomaly detection techniques, which do not account for highly skewed edge and degree statistics. To address this problem we introduce HYPA, a novel framework for the unsupervised detection of anomalies in large corpora of variable-length temporal paths in a graph. HYPA provides an efficient analytical method to detect paths with anomalous frequencies that result from nodes being traversed in unexpected chronological order.

show abstract

Section: Related Workmentioning

confidence: 99%

HYPA: Efficient Detection of Path Anomalies in Time Series Data on Networks

LaRock

Nanumyan

Scholtes

et al. 2020

Proceedings of the 2020 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

show abstract

“…The anomaly scores are finally determined based on the residuals. Inside this category, we can include anomaly detection techniques based on traditional time series forecasting models such as Vector Auto-Regressive (VAR) [38,39] and Autoregressive Integrated Moving Average (ARIMA) [40,41]. Also, RNN have been used as regression models and will be covered later on in a specific section of the survey.…”

Section: Regression Model-basedmentioning

confidence: 99%

“…Melnyk et al [38] propose an unsupervised model-based framework adapted to online anomaly detection where each flight is represented as a Vector AutoRegressive eXogenous model (VARX) model [115]. The key step in the approach is to compute a distance matrix between flights defined in terms of residuals of modeling one flight's data using another flight's VARX model.…”

Section: Statistical-based Approachesmentioning

confidence: 99%

Recent Advances in Anomaly Detection Methods Applied to Aviation

2019

View full text Add to dashboard Cite

Anomaly detection is an active area of research with numerous methods and applications. This survey reviews the state-of-the-art of data-driven anomaly detection techniques and their application to the aviation domain. After a brief introduction to the main traditional data-driven methods for anomaly detection, we review the recent advances in the area of neural networks, deep learning and temporal-logic based learning. In particular, we cover unsupervised techniques applicable to time series data because of their relevance to the aviation domain, where the lack of labeled data is the most usual case, and the nature of flight trajectories and sensor data is sequential, or temporal. The advantages and disadvantages of each method are presented in terms of computational efficiency and detection efficacy. The second part of the survey explores the application of anomaly detection techniques to aviation and their contributions to the improvement of the safety and performance of flight operations and aviation systems. As far as we know, some of the presented methods have not yet found an application in the aviation domain. We review applications ranging from the identification of significant operational events in air traffic operations to the prediction of potential aviation system failures for predictive maintenance.

show abstract

Section: Regression Model-basedmentioning

confidence: 99%

“…In this framework, aircraft flying in the same airspace are represented as images and the ConvLSTM-AE model is used to detect anomalies in the sequences of images leading to anomalous ADS-B location reports.4.1.4. Statistical-based approachesMelnyk et al[38] propose an unsupervised model-based framework adapted to online anomaly detection where each flight is represented as a Vector AutoRegressive eXogenous model (VARX) model[115]. The key step in the approach is to compute a distance matrix between flights defined in terms of residuals of modeling one flight's data using another flight's VARX model.…”

mentioning

confidence: 99%

Recent Advances in Anomaly Detection Methods applied to Aviation

Basora¹,

Olivé²,

Dubot³

2019

Preprint

View full text Add to dashboard Cite

Anomaly detection is an active area of research with numerous methods and applications. This survey reviews the state-of-the-art of data-driven anomaly detection techniques and their application to the the aviation domain. After a brief introduction to the main traditional data-driven methods for anomaly detection, we review the recent advances in the area of neural networks, deep learning and temporal-logic based learning. We cover especially unsupervised techniques applicable to time series data because of their relevance to the aviation domain, where the lack of labeled data is the most usual case, and the nature of flight trajectories and sensor data is sequential, or temporal. The advantages and disadvantages of each method are presented in terms of computational efficiency and detection efficacy. The second part of the survey explores the application of anomaly detection techniques to aviation and their contributions to the improvement of the safety and performance of flight operations and aviation systems. As far as we know, some of the presented methods have not yet found an application in the aviation domain. We review applications ranging from the identification of significant operational events in air traffic operations to the prediction of potential aviation system failures for predictive maintenance.

show abstract

Vector Autoregressive Model-Based Anomaly Detection in Aviation Systems

Cited by 37 publications

References 28 publications

HYPA: Efficient Detection of Path Anomalies in Time Series Data on Networks

HYPA: Efficient Detection of Path Anomalies in Time Series Data on Networks

Recent Advances in Anomaly Detection Methods Applied to Aviation

Recent Advances in Anomaly Detection Methods applied to Aviation

Contact Info

Product

Resources

About