Abstract
OECD Benchmark data distributed for the SMORN VI Specialists Meeting in Reactor Noise were investigated for anomaly detection in artificially generated reactor noise benchmark analysis. It was observed that statistical features extracted from covariance matrix of frequency components are very sensitive in terms of the anomaly detection level. It is possible to create well defined alarm levels. (R.P.) 5 refs.; 23 figs.; 1 tab.
Srinivasan, G S;
[1]
Krinizs, K;
[2]
Por, G
[3]
- Indira Gandhi Centre for Atomic Research, Kalpakkam (India)
- Hungarian Academy of Sciences, Budapest (Hungary). Central Research Inst. for Physics
- Budapesti Mueszaki Egyetem (Hungary). Inst. for Nuclear Techniques
Citation Formats
Srinivasan, G S, Krinizs, K, and Por, G.
Anomaly detection in OECD Benchmark data using co-variance methods.
Hungary: N. p.,
1993.
Web.
Srinivasan, G S, Krinizs, K, & Por, G.
Anomaly detection in OECD Benchmark data using co-variance methods.
Hungary.
Srinivasan, G S, Krinizs, K, and Por, G.
1993.
"Anomaly detection in OECD Benchmark data using co-variance methods."
Hungary.
@misc{etde_10112981,
title = {Anomaly detection in OECD Benchmark data using co-variance methods}
author = {Srinivasan, G S, Krinizs, K, and Por, G}
abstractNote = {OECD Benchmark data distributed for the SMORN VI Specialists Meeting in Reactor Noise were investigated for anomaly detection in artificially generated reactor noise benchmark analysis. It was observed that statistical features extracted from covariance matrix of frequency components are very sensitive in terms of the anomaly detection level. It is possible to create well defined alarm levels. (R.P.) 5 refs.; 23 figs.; 1 tab.}
place = {Hungary}
year = {1993}
month = {Feb}
}
title = {Anomaly detection in OECD Benchmark data using co-variance methods}
author = {Srinivasan, G S, Krinizs, K, and Por, G}
abstractNote = {OECD Benchmark data distributed for the SMORN VI Specialists Meeting in Reactor Noise were investigated for anomaly detection in artificially generated reactor noise benchmark analysis. It was observed that statistical features extracted from covariance matrix of frequency components are very sensitive in terms of the anomaly detection level. It is possible to create well defined alarm levels. (R.P.) 5 refs.; 23 figs.; 1 tab.}
place = {Hungary}
year = {1993}
month = {Feb}
}