PCA-based network Traffic anomaly detection

Ding, Meimei; Tian, Hui

doi:10.1109/tst.2016.7590319

Cited by 40 publications

(3 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Note that the particular power-fullness of an auto-encoder, a nonlinear PCA, as a data compressor is not desirable herein since auto-encoders compress all patterns including abnormal ones. Although [38] claimed that the PCA-based models are stable with respect to their parameters, such as the number of principal components k spanning the subspace or the cutoff level, Ref. [36] found instead that PCA-based anomaly detection is sensitive to these parameters and to the amplitude of the anomalies.…”

Section: Conflicts Of Interestmentioning

confidence: 99%

Anomaly Detection in Financial Time Series by Principal Component Analysis and Neural Networks

Crépey

Noureddine²,

Madhar

et al. 2022

Algorithms

View full text Add to dashboard Cite

A major concern when dealing with financial time series involving a wide variety of market risk factors is the presence of anomalies. These induce a miscalibration of the models used to quantify and manage risk, resulting in potential erroneous risk measures. We propose an approach that aims to improve anomaly detection in financial time series, overcoming most of the inherent difficulties. Valuable features are extracted from the time series by compressing and reconstructing the data through principal component analysis. We then define an anomaly score using a feedforward neural network. A time series is considered to be contaminated when its anomaly score exceeds a given cutoff value. This cutoff value is not a hand-set parameter but rather is calibrated as a neural network parameter throughout the minimization of a customized loss function. The efficiency of the proposed approach compared to several well-known anomaly detection algorithms is numerically demonstrated on both synthetic and real data sets, with high and stable performance being achieved with the PCA NN approach. We show that value-at-risk estimation errors are reduced when the proposed anomaly detection model is used with a basic imputation approach to correct the anomaly.

show abstract

Section: Conflicts Of Interestmentioning

confidence: 99%

Anomaly Detection in Financial Time Series by Principal Component Analysis and Neural Networks

Crépey

Noureddine²,

Madhar

et al. 2022

Algorithms

View full text Add to dashboard Cite

show abstract

“…Their method was evaluated against other techniques and revealed that it was more accurate at detecting anomalies. M. Ding et al [11] proposed a method that uses the PCA model to reduce the dimensionality of the data. It then uses the residuals of the model to detect anomalies.…”

Section: Literature Reviewmentioning

confidence: 99%

Anomaly Detection in Network Traffic using Machine Learning and Deep Learning Techniques

Rana¹

2019

TURCOMAT

View full text Add to dashboard Cite

Due to the rise of sophisticated cyberattacks, network security has become an increasingly important field. One of the most common threats to the security of networks is network anomalies, which can cause system malfunctions and prevent them from working properly. Detecting such anomalies is very important to ensure the continued operation of the network. Deep learning and machine learning algorithms have demonstrated their ability to detect network anomalies, but their effectiveness is still not widely known. This paper presents an evaluation of the performance of three algorithms against the KDD-NSL dataset. This study aims to provide a comprehensive analysis of the various techniques used in deep learning and machine learning to detect network anomalies. It will also help improve the security of networks. The paper presents an evaluation of the performance of three algorithms against the KDD-NSL dataset. The three algorithms are the Support Vector Machine, the Random Forest, and the Artificial Neural Network. They will be compared with their accuracy, recall, and F1-score. The study also explores the impact of the algorithm's feature selection on its performance. The findings of the investigation will be used to inform the development of new techniques that can be utilized to enhance the security of networks. The KDD NSL dataset provides an ideal opportunity to analyze the performance of various algorithms for detecting network anomalies.

show abstract

“…However this approach depends both on the anomaly detection technique and the properties of the feature space. For example clustering based methods [10] [13] [19] require that the anomalies do not aggregate into clusters; nearest neighbour and density based methods [6] require that the anomalies do not form dense regions in the feature space; spectral methods [7] [3] assume that a projection into a different space exists such that normal and anomalous points can be clearly distinguished. Another approach consists in training a model to predict the signal in the future and then compare the predicted and observed signals to detect anomalies, like in [14].…”

Section: Introductionmentioning

confidence: 99%

A GAN Approach for Anomaly Detection in Spacecraft Telemetries

Ciancarelli

Magistris

Cognetta

et al. 2022

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

In spacecraft health management a large number of time series is acquired and used for on-board units surveillance and for historical data analysis. The early detection of abnormal behaviors in telemetry data can prevent failures in the spacecraft equipment. In this paper we present an advanced monitoring system that was carried out in partnership with Thales Alenia Space Italia S.p.A, a leading industry in the field of spacecraft manufacturing. In particular, we developed an anomaly detection algorithm based on Generative Adversarial Networks, that thanks to their ability to model arbitrary distributions in high dimensional spaces, allow to capture complex anomalies avoiding the burden of hand crafted feature extraction. We applied this method to detect anomalies in telemetry data collected from a simulator of a Low Earth Orbit satellite. One of the strengths of the proposed approach is that it does not require any previous knowledge on the signal. This is particular useful in the context of anomaly detection where we do not have a model of the anomaly. Hence the only assumption we made is that an anomaly is a pattern that lives in a lower probability region of the data space.

show abstract

PCA-based network Traffic anomaly detection

Cited by 40 publications

References 14 publications

Anomaly Detection in Financial Time Series by Principal Component Analysis and Neural Networks

Anomaly Detection in Financial Time Series by Principal Component Analysis and Neural Networks

Anomaly Detection in Network Traffic using Machine Learning and Deep Learning Techniques

A GAN Approach for Anomaly Detection in Spacecraft Telemetries

Contact Info

Product

Resources

About