ACM Transactions on Computer Systems · 2002 · 58 citations · 30 references
ComputingComputer ScienceComputer SystemsDynamical SystemsStochastic ProcessesTime Series AnalysisAnomaly DetectionAnomaly Detection In Time SeriesHigh-frequency Data AnalysisOutlier DetectionComparative Data AnalysisCorrelationCorrelation LengthDimensionless QuantitiesPeriodicityTime Series Outlier DetectionTransaction Time
A comparative analysis of transaction time-series is made, for light to moderately loaded hosts, motivated by the problem of anomaly detection in computers. Criteria for measuring the statistical state of hosts are examined. Applying a scaling transformation to the measured data, it is found that the distribution of fluctuations about the mean is closely approximated by a steady-state, maximum-entropy distribution, modulated by a periodic variation. The shape of the distribution, under these conditions, depends on the dimensionless ratio of the daily/weekly periodicity and the correlation length of the data. These values are persistent or even invariant. We investigate the limits of these conclusions, and how they might be applied in anomaly detection.
30
<i>The Mathematical Theory of Communication</i>
Claude E. Shannon, Warren Weaver, Norbert Wiener · Physics Today · 1950 · 9.9K citations
Wide area traffic: the failure of Poisson modeling
Vern Paxson, Sally Floyd · IEEE/ACM Transactions on Networking · 1995 · 3.7K citations · Full text
Cambridge University Press · 2021 · 1.2K citations
Stephanie Forrest, Steven Hofmeyr, Anil Somayaji · Communications of the ACM · 1997 · 853 citations · Full text