Expert-Based Fusion Algorithm of an Ensemble of Anomaly Detection Algorithms - Technologies and Applications of Artificial Intelligence - page 122

6Conluon
In this paper we consider the Fusion of multiple anomaly detection algorithm.
The motivation for this fusion process has evolved due to the widespread belief
that even though none of the existing ADAs achieves perfect classification, the
combination of multiple ADAs may create a superior outlier detection algorithm
as has been achieved in the classification and clustering domains. In this paper we
describe the expertise based fusion algorithm we developed. This algorithm may
be classified as a semi-supervised method. To evaluate the performance of the
proposed method with the benchmark method that exists in the literature, we
limited the study to the case of a single type of outlier. For this case we showed
that by using the /union voting method, we can overcome the normalization
problem which is one of the critical parts in any fusion process. We do so by
using ranking instead of actual scores. Thus we demonstrated that our method
outperforms the benchmark method from the literature.
References
1. Chandola, V., Banerjee, A., Kumar, V.: Outlier detection: A survey. ACM Com-
puting Surveys (2007) (to appear)
2. Petrovskiy, M.I.: Outlier detection algorithms in data mining systems. Program-
ming and Computer Software 29(4), 228-237 (2003)
3. Zhang, L., Leung, H., Chan, K.C.C.: Information fusion based smart home control
system and its application. IEEE Transactions on Consumer Electronics 54(3),
1157-1165 (2008)
4. Ahmed, M., Pottie, G.: Fusion in the context of information theory. Distributed
Sensor Networks, 419-436 (2005)
5. Jeon, B., Landgrebe, D.A.: Decision fusion approach for multitemporal classifi-
cation. IEEE Transactions on Geoscience and Remote Sensing 37(3), 1227-1233
(1999)
6. Schubert, E., et al.: On Evaluation of Outlier Rankings and Outlier Scores. In:
SDM (2012)
7. Dietterich, T.G.: Ensemble methods in machine learning. Multiple classifier sys-
tems, pp. 1-15. Springer, Heidelberg (2000)
8. Tan, A.C., Gilbert, D.: Ensemble machine learning on gene expression data for
cancer classification (2003)
9. Balke, W.-T., Kießing, W.: Optimizing multi-feature queries for image databases.
In: Proc. of the Intern. Conf. on Very Large Databases (2000)
10. Asuncion, A., Newman, D.: UCI machine learning repository (2007)
11. Kriegel, H.-P., Kroger, P., Schubert, E., Zimek, A.: Interpreting and unifying outlier
scores. In: Proc. SDM, pp. 13-24 (2011)
12. Lazarevic, A., Kumar, V.: Feature bagging for outlier detection. In: Proc. KDD,
pp. 157-166 (2005)
13. Nguyen, H.V., Ang, H.H., Gopalkrishnan, V.: Mining outliers with ensemble of
heterogeneous detectors on random subspaces. In: Kitagawa, H., Ishikawa, Y.,
Li, Q., Watanabe, C. (eds.) DASFAA 2010. LNCS, vol. 5981, pp. 368-383. Springer,
Heidelberg (2010)