ABSTRACT
One class classification is an effective way to perform anomaly detection or outlier detection. Previous work has empirically shown that complexity in the background (also known as target) class degrades the performance of one-class classifiers, and one source of data complexity is the presence of subconcepts in that class. Learning over subconcepts individually can mitigate the effects of domain complexity and improve one class classification performance. Unless a clustering could be derived from domain knowledge, the approach used to search for subconcepts is unsupervised clustering. However, in some cases, the examples belonging to some of the clusters may be too scattered, while, in others, clusters may be affected by the small disjunct problem where some clusters comprise only of a few examples that cannot be used to train a robust classifier. To handle these problems, this paper presents a method to update the clusters obtained by the clustering process prior to learning over them. In addition, it introduces the c-Nearest-Cluster (c-NC) method for combining the individual classifiers derived from each subconcept. Our experiments on classical outlier detection datasets and on cyber security datasets show that our method can improve upon the classification performance obtained by the recently proposed subconcept based one class classification method.
- C. Bellinger, S. Sharma, O. R. Zaiane, et al. Sampling a Longer Life: Binary versus One-class classification revisited, Proceedings of Machine Learning Research, 2017, 74:64--78.Google Scholar
- C. Bellinger, S. Sharma, N. Japkowicz. One-Class versus Binary Classification: Which and When? 11th International Conference on Machine Learning and Applications (ICMLA), 2012:102--106.Google Scholar
- S. Sharma, A. Somayaji, N. Japkowicz. Learning over subconcepts: Strategies for 1-class classification. Computational Intelligence, 2018, 34(2): 440--467.Google ScholarCross Ref
- F. T. Liu, K. M. Ting, Z. H. Zhou. Isolation forest. In 2008 Eighth IEEE International Conference on Data Mining, 2008, pp. 413--422.Google ScholarDigital Library
- C. Bellinger, S. Sharma, N. Japkowicz. One-class classification--From theory to practice: A case-study in radioactive threat detection. Expert Systems with Applications, 2018, 108: 223--232.Google ScholarDigital Library
- J. Yu, S. Kang. Clustering-based proxy measure for optimizing one-class classifiers. Pattern Recognition Letters, 2019, 117: 37--44.Google ScholarCross Ref
- H. Zuo, O. Wu, W. Hu, et al. Recognition of blue movies by fusion of audio and video, In IEEE International Conference on Multimedia and Expo, 2008, pp. 37--40.Google Scholar
- B. Krawczyk, M. Woźniak. Dynamic classifier selection for one-class classification. Knowledge-Based Systems, 2016, 107, 43--53.Google ScholarDigital Library
- D.M.J. Tax, R.P.W. Duin, Support vector data description, Machine Learning, 2004, 54 (1): 45--66.Google ScholarDigital Library
- B. Chen, A. Feng, S. Chen, B. Li, One-cluster clustering based data description, Jisuanji Xuebao/Chinese J. Comput. 2007, 30(8): 1325--1332.Google Scholar
- L. Swersky, H. O. Marques, J. Sander, et al. On the evaluation of outlier detection and one-class classification methods. In IEEE International Conference on Data Science and Advanced Analytics (DSAA), 2016: 1--10.Google Scholar
- E. R. de Faria, A. C. P. de Leon Ferreira, J. Gama. MINAS: multiclass learning algorithm for novelty detection in data streams. Data mining and knowledge discovery, 2016, 30(3): 640--680.Google Scholar
- ODDS library, http://odds.cs.stonybrook.eduGoogle Scholar
- G. Creech, J. Hu. Generation of a new IDS test dataset: Time to retire the KDD collection. In IEEE Wireless Communications and Networking Conference (WCNC), 2013, pp. 4487--4492.Google ScholarCross Ref
- W. Haider, J. Hu, J. Slay, et al. Generating realistic intrusion detection system dataset based on fuzzy qualitative modeling. Journal of Network and Computer Applications, 2017, 87, C: 185--192.Google ScholarDigital Library
- M. Xie, J. Hu. Evaluating host-based anomaly detection systems: A preliminary analysis of adfa-ld. In 6th International Congress on Image and Signal Processing (CISP), 2013, pp. 1711--1716.Google ScholarCross Ref
- K. Xu, F. Wang. Behavioral Graph Analysis of Internet Applications. In Global Telecommunications Conference. 2015, pp. 1763--1768.Google Scholar
Index Terms
- Subconcept Based One Class Classification Method with Cluster Updating
Recommendations
One class random forests
One class classification is a binary classification task for which only one class of samples is available for learning. In some preliminary works, we have proposed One Class Random Forests (OCRF), a method based on a random forest algorithm and an ...
Classification by cluster analysis: a new meta-learning based approach
MCS'11: Proceedings of the 10th international conference on Multiple classifier systemsCombination of multiple classifiers, commonly referred to as an classifier ensemble, has previously demonstrated the ability to improve classification accuracy in many application domains. One popular approach to building such a combination of ...
Cluster-based one-class ensemble for classification problems in information retrieval
SIGIR '12: Proceedings of the 35th international ACM SIGIR conference on Research and development in information retrievalA number of relevant information retrieval classification problems are one-class classification problems at heart. I.e., labeled data is only available for one class, the so-called target class, and common discrimination-based classification approaches, ...
Comments