TY - GEN
T1 - Contributor diagnostics for anomaly detection
AU - Borisov, Alexander
AU - Runger, George
AU - Tuv, Eugene
PY - 2009/11/27
Y1 - 2009/11/27
N2 - Anomaly detection in data streams requires a signal of an unusual event, but an actionable response requires diagnostics. Consequently, an important task is to isolate to the few key attributes that contribute to the signal from among a large collection. We introduce this contributor problem to the machine learning community and present a solution for monitoring in modern systems (with nonlinear reference conditions, high dimensions, categorical attributes, missing data, and so forth). The objective is to identify attributes that contribute to a signal, for both individual and multiple anomalies, or from several anomaly groups. Although related to the feature selection problem, the extreme sparseness of anomalies leads to scores that are designed specifically for the contributors problem. Statistical criteria are provided to quantitatively address decision rules and false alarms and the method can be computed quickly. Comparisons are made to traditional contribution plots.
AB - Anomaly detection in data streams requires a signal of an unusual event, but an actionable response requires diagnostics. Consequently, an important task is to isolate to the few key attributes that contribute to the signal from among a large collection. We introduce this contributor problem to the machine learning community and present a solution for monitoring in modern systems (with nonlinear reference conditions, high dimensions, categorical attributes, missing data, and so forth). The objective is to identify attributes that contribute to a signal, for both individual and multiple anomalies, or from several anomaly groups. Although related to the feature selection problem, the extreme sparseness of anomalies leads to scores that are designed specifically for the contributors problem. Statistical criteria are provided to quantitatively address decision rules and false alarms and the method can be computed quickly. Comparisons are made to traditional contribution plots.
UR - http://www.scopus.com/inward/record.url?scp=70450204281&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70450204281&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-04277-5_95
DO - 10.1007/978-3-642-04277-5_95
M3 - Conference contribution
AN - SCOPUS:70450204281
SN - 3642042767
SN - 9783642042768
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 944
EP - 953
BT - Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings
T2 - 19th International Conference on Artificial Neural Networks, ICANN 2009
Y2 - 14 September 2009 through 17 September 2009
ER -