Filters








6,622 Hits in 6.3 sec

Online false discovery rate control for anomaly detection in time series [article]

Quentin Rebjock, Barış Kurt, Tim Januschowski, Laurent Callot
2021 arXiv   pre-print
This article proposes novel rules for false discovery rate control (FDRC) geared towards online anomaly detection in time series.  ...  rare (typical in anomaly detection) and the test statistics are serially dependent (typical in time series).  ...  Online FDRC methods are appealing for anomaly detection problems as controlling the false discovery rate is equivalent to maximizing recall given a lower bound on precision.  ... 
arXiv:2112.03196v1 fatcat:tou667s475b5zb4jcrifea2fqe

Scalable Architecture for Anomaly Detection and Visualization in Power Generating Assets

Paras Jain, Chirag Tailor, Sam Ford, Liexiao Ding, Michael Phillips, Fang Liu, Nagi Gebraeel, Duen Horng Chau
2017 2017 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)  
To reduce false alarm rates, we leverage the False Discovery Rate (FDR) algorithm which significantly reduces the number of false alarms.  ...  ; (2) prevalence of false alarms generated by anomaly detection algorithms resulting in unnecessary downtime and maintenance; and (3) lack of an integrated visualization that helps users understand and  ...  for their generous 200-machine donation. We thank Will Powell on hardware support for our system.  ... 
doi:10.1109/ipdpsw.2017.77 dblp:conf/ipps/JainTFDPLGC17 fatcat:g7lzys3mbncqrcxeszatcrmzdq

Scalable Architecture for Anomaly Detection and Visualization in Power Generating Assets [article]

Paras Jain, Chirag Tailor, Sam Ford, Liexiao Ding, Michael Phillips, Fang Liu, Nagi Gebraeel, Duen Horng Chau
2017 arXiv   pre-print
To reduce false alarm rates, we leverage the False Discovery Rate (FDR) algorithm which significantly reduces the number of false alarms.  ...  ; (2) prevalence of false alarms generated by anomaly detection algorithms resulting in unnecessary downtime and maintenance; and (3) lack of an integrated visualization that helps users understand and  ...  for their generous 200-machine donation. We thank Will Powell on hardware support for our system.  ... 
arXiv:1701.07500v1 fatcat:y32buwiuk5elzmppdfek6nxbrq

Cost-Effective Bad Synchrophasor Data Detection Based on Unsupervised Time Series Data Analytics [article]

Lipeng Zhu, David J. Hill
2020 arXiv   pre-print
To make the whole approach competent in processing online streaming PMU data, an efficient strategy for accelerating STNN discovery is tactfully designed.  ...  Considering the imperative need for filtering out potential bad data, this paper develops a simple yet efficient online bad PMU data detection approach by exploring spatial-temporal correlations.  ...  Both Remark 1: Misdetection rate is the ratio of falsely dismissed anomalous instances (with bad data), while false alarm rate stands for what percentage of normal instances is falsely labeled as anomalies  ... 
arXiv:2005.01060v1 fatcat:imamzpmufzbwfnc7ylrfnbfdje

A Comprehensive Survey of Time Series Anomaly Detection in Online Social Network Data

Md Rafiqul, Naznin Sultana, Mohammad Ali, Prohollad Chandra, Bushra Rahman
2017 International Journal of Computer Applications  
We provide some applications, challenging issues and existing methods for time series anomaly detection.  ...  In the field of data mining, the social network is one of the complex systems that poses significant challenges in this area. Time series anomaly detection is one of the critical applications.  ...  Table 2 : 2 List of time series anomaly detection methodsUsed for identifing unusual time series in a large collections of time series.  ... 
doi:10.5120/ijca2017915989 fatcat:2h2darlw3jd7jedvb723p42vtm

Finding Needle in a Million Metrics: Anomaly Detection in a Large-scale Computational Advertising Platform [article]

Bowen Zhou, Shahriar Shariat
2016 arXiv   pre-print
In this paper we describe the mechanism that we invented for recovering the representative metrics and detecting the change in their behavior.  ...  We show that this mechanism is able to detect the possible problems in time by describing some incident cases.  ...  In fact we would prefer a to have a few false alarms in the interest of a higher discovery rate. • The alerts should be triggered within reasonable amount of time, typically a couple of hours since the  ... 
arXiv:1602.07057v1 fatcat:45hi744edva3xp26piopgg5gjm

Exponentially Weighted Ellipsoidal Model for Anomaly Detection

M. Moshtaghi, S. M. Erfani, C. Leckie, J.C. Bezdek
2017 International Journal of Intelligent Systems  
Recently, an online efficient anomaly detection technique called Iterative Data Capture Anomaly Detection has been proposed for environmental sensing and monitoring applications.  ...  Despite the real-time nature of the data collected in the IoT and limited memory and computational resources, most of the current data modeling approaches for the IoT involve batch training.  ...  S −1 k+1,λ using ( 14 ); Algorithm 1: ewIDCAD algorithm for input x k+1 In this section, we describe two prominent online approaches for anomaly detection in multivariate time-series.  ... 
doi:10.1002/int.21875 fatcat:hvwwvbd4djafbedwdabpohmo2a

Adaptive anomaly detection with evolving connectionist systems

Yihua Liao, V. Rao Vemuri, Alejandro Pasos
2007 Journal of Network and Computer Applications  
Anomaly detection holds great potential for detecting previously unknown attacks.  ...  Neural Networks (EFuNN), can significantly reduce the false alarm rate while the attack detection rate remains high. r  ...  This work is supported in part by the AFOSR Grant F49620-01-1-0327 to the Center for Digital Security of the University of California, Davis.  ... 
doi:10.1016/j.jnca.2005.08.005 fatcat:4woc5j3ssbat7gksaa7mqpldhe

Structure-Adaptive Sequential Testing for Online False Discovery Rate Control [article]

Bowen Gang, Wenguang Sun, Weinan Wang
2020 arXiv   pre-print
This work develops a new class of structure–adaptive sequential testing (SAST) rules for online false discover rate (FDR) control.  ...  Consider the online testing of a stream of hypotheses where a real–time decision must be made before the next data point arrives. The error rate is required to be controlled at all decision points.  ...  We discuss two applications, respectively for anomaly detection in large-scale time series data and genotype discovery under the QPD framework.  ... 
arXiv:2003.00113v1 fatcat:vyliaik3gvb6dcwuhfdrstuadi

FChain: Toward Black-Box Online Fault Localization for Cloud Systems

Hiep Nguyen, Zhiming Shen, Yongmin Tan, Xiaohui Gu
2013 2013 IEEE 33rd International Conference on Distributed Computing Systems  
In this paper, we present a black-box online fault localization system called FChain that can pinpoint faulty components immediately after a performance anomaly is detected.  ...  FChain performs runtime validation to further filter out false alarms.  ...  The authors would like to thank the anonymous reviewers for their insightful comments.  ... 
doi:10.1109/icdcs.2013.26 dblp:conf/icdcs/NguyenSTG13 fatcat:cyq3ffdutvefhgn2kw6vtwk24u

Categorizing and comparing psychophysical detection strategies based on biomechanical responses to short postural perturbations

Viprali V Bhatkar, Joseph D Skufca, Rakesh B Pilkar, Christopher M Storey, Charles J Robinson
2010 BioMedical Engineering OnLine  
A time-series-bitmap approach was used to identify anomalies in interval 1 (a 1 ) and interval 2 (a 2 ) that were present in the resultant APCoP signal. If a 1 > a 2 then R B = Interval 1.  ...  A fundamental unsolved problem in psychophysical detection experiments is in discriminating guesses from the correct responses.  ...  Acknowledgements We thank psychophysicists Robert Carlson and Alan Searleman for their very constructive comments.  ... 
doi:10.1186/1475-925x-9-58 pmid:20932297 pmcid:PMC2959020 fatcat:rxuit7w6tfa3xnwnhpeyrqi2me

Anomaly detection using clustering for ad hoc networks -behavioral approach-

Belacel Madani, B. Messabih
2012 Computer Engineering and Applications Journal  
The anomaly detection algorithms have the advantagebecause they can detect new types of attacks (zero-day attacks).In this paper, wepresent a Intrusion Detection System clustering-based (ID-Cluster) that  ...  In MANETs, it isdifficult to detect malicious nodes because the network topology constantly changesdue to node mobility.  ...  The routing control packets in a given time interval are not sufficient for the intrusion detection.  ... 
doi:10.18495/comengapp.v1i1.6 fatcat:qmml3gzlsreqjhprreqrrdvsla

Conformal k-NN Anomaly Detector for Univariate Data Streams [article]

Vladislav Ishimtsev, Ivan Nazarov, Alexander Bernstein, Evgeny Burnaev
2017 arXiv   pre-print
In this paper we consider a model-free anomaly detection method for univariate time-series which adapts to non-stationarity in the data stream and provides probabilistic abnormality scores based on the  ...  Anomalies in time-series data give essential and often actionable information in many applications.  ...  The number of false negatives is the number of anomaly windows in the time-series, with no true positive detections. True negatives are not used in scoring.  ... 
arXiv:1706.03412v1 fatcat:zgzwdous6rg6bputsvnowwbno4

Collecting Labels for Rare Anomalies via Direct Human Feedback—An Industrial Application Study

Christian Reich, Ahmad Mansour, Kristof Van Laerhoven
2019 Informatics  
For the detection of anomalies in industrial settings, sensor units have been introduced to predict and classify such anomalous events, but these critically rely on annotated data for training.  ...  A prototype for visualization and in situ annotation of sensor signals is developed with embedded unsupervised anomaly detection algorithms to propose signals for annotation and which allows the operators  ...  Depending on the chosen anomaly detection algorithm, this dominance of spurious outliers typically results in either a high false positive rate (FPR) or false negative rate (FNR).  ... 
doi:10.3390/informatics6030038 fatcat:uzbzl7rspjfnzaalyshkeqb53u

Centralized and Distributed Intrusion Detection for Resource Constrained Wireless SDN Networks [article]

Gustavo A. Nunez Segura, Arsenia Chorti, Cintia Borges Margi
2021 arXiv   pre-print
There are proposals in the literature to detect DoS in wireless SDN networks, however, not without shortcomings: there is little focus on resource constraints, high detection rates have been reported only  ...  The results show detection rates exceeding 96% in networks of 36 and 100 nodes and identification of the type of the attack with a probability exceeding 0.89 when using the centralized approach.  ...  In this work, we employed a CUSUM based algorithm to detect changes in the mean value of control overhead and data packets delivery rate time series.  ... 
arXiv:2103.01262v1 fatcat:ffuixybxqndmpooder4lyh52ae
« Previous Showing results 1 — 15 out of 6,622 results