Abstract | ||
---|---|---|
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. |
Year | DOI | Venue |
---|---|---|
2002 | 10.1145/507052.507054 | ACM Trans. Comput. Syst. |
Keywords | Field | DocType |
anomaly detection,statistical state,dimensionless ratio,periodic variation,measured data,correlation length,scaling transformation,maximum-entropy distribution,comparative analysis,statistical mechanics,transaction time-series,system normality,time series,measurement system,maximum entropy,steady state | Normality,Statistical physics,Anomaly detection,Statistical mechanics,Computer science,Correlation function (statistical mechanics),Real-time computing,Invariant (mathematics),Scaling,Periodic graph (geometry),Dimensionless quantity | Journal |
Volume | Issue | ISSN |
20 | 2 | 0734-2071 |
Citations | PageRank | References |
26 | 1.77 | 26 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Mark Burgess | 1 | 203 | 22.41 |
Hårek Haugerud | 2 | 39 | 5.90 |
Sigmund Straumsnes | 3 | 29 | 2.21 |
Trond Reitan | 4 | 31 | 3.15 |