Physics-Based Anomaly Detection Defined on Manifold Space
作者： Hao Huang , Hong Qin , Shinjae Yoo , Dantong Yu
DOI: 10.1145/2641574
关键词:
摘要: Current popular anomaly detection algorithms are capable of detecting global anomalies but often fail to distinguish local from normal instances. Inspired by contemporary physics theory (i.e., heat diffusion and quantum mechanics), we propose two unsupervised algorithms. Building on the embedding manifold derived diffusion, devise Local Anomaly Descriptor (LAD), which faithfully reveals intrinsic neighborhood density. It uses a scale-dependent umbrella operator bridge properties, makes LAD more informative within an adaptive scope neighborhood. To offer stability density measurement scaling parameter tuning, formulate Fermi Density (FDD), measures probability fermion particle being at specific location. By choosing stable energy distribution function, FDD steadily distinguishes instances with any setting. further enhance efficacy our proposed algorithms, explore utility anisotropic Gaussian kernel (AGK), offers better manifold-aware affinity information. We also quantify examine effect different Laplacian normalizations for detection. Comprehensive experiments both synthetic benchmark datasets verify that outperform existing
acm.org
osti.gov
sci-hub.se 参考文章(56)
R. Kagan, E. M. Heenan, F. Handa, H. Hayashii, H. Aihara, H. Aihara, J. Kaneko, H. Ishino, A. Ishikawa, A. Ishikawa, H. C. Huang, Y. Chao, Y. Chao, Y. Chao, J. S. Kang, T. Hara, T. Hara, Byoung Sup Ahn, Byoung Sup Ahn, Byoung Sup Ahn, T. Kawasaki, T. Kawasaki, T. Kawasaki, N. Katayama, N. Katayama, N. Katayama, S. Kobayashi, S. Kobayashi, S. Kobayashi, Heejong Kim, Heejong Kim, Heejong Kim, R. Abe, K. Abe, R. Guo, L. Y. Dong, H. K. Jang, A. Kuzmin, A. Kuzmin, A. Kuzmin, J. H. Kang, D. Liventsev, D. Liventsev, D. Liventsev, H. J. Kim, H. J. Kim, H. J. Kim, S. H. Lee, S. H. Lee, S. H. Lee, J. Haba, Y. Choi, R. Chistov, R. Chistov, R. Chistov, P. Krokovny, P. Krokovny, P. Krokovny, J. S. Lange, J. S. Lange, J. S. Lange, M. Hazumi, G. Leder, G. Leder, G. Leder, A. Bozek, K. Gotow, K. Kinoshita, K. Kinoshita, K. Kinoshita, H. Iwasaki, H. Iwasaki, H. Hirano, N. Gabyshev, P. Chang, P. Chang, P. Chang, A. Gordon, I. Higuchi, I. Higuchi, Y. Enari, Y. Enari, M. Fukushima, M. Fukushima, W. S. Hou, B. C.K. Casey, B. C.K. Casey, H. Kichimi, H. Kichimi, H. Kichimi, A. Bondar, R. Kulasiri, R. Kulasiri, R. Kulasiri, J. Dragic, D. W. Kim, D. W. Kim, D. W. Kim, E. Banas, E. Banas, E. Banas, K. Hara, T. Hojo, T. Aushev, T. Aushev, T. Aushev, B. G. Cheon, B. G. Cheon, B. G. Cheon, K. Inami, K. Inami, R. Itoh, R. Itoh, H. Fujii, N. Kawamura, N. Kawamura, N. Kawamura, T. Hokuue, S. Eidelman, S. Eidelman, S. Eidelman, H. Konishi, H. Konishi, H. Konishi, A. Garmash, H. Hamasaki, S. Behari, S. Behari, S. Behari, Y. Igarashi, K. Hoshina, K. Hoshina, A. Drutskoy, T. Aso, T. Aso, T. Aso, H. Ikeda, H. Ikeda, H. Ikeda, V. Aulchenko, V. Aulchenko, V. Aulchenko, R. S. Lu, R. S. Lu, R. S. Lu, Y. Iwasaki, C. Fukunaga, S. Kumar, S. Kumar, S. Kumar, P. K. Behera, P. K. Behera, P. K. Behera, Y. Asano, Y. Asano, Y. Asano, Y. J. Kwon, Y. J. Kwon, Y. J. Kwon, D. J. Jackson, A. M. Bakich, A. M. Bakich, S. K. Kim, S. K. Kim, S. K. Kim, Y. Hoshi, H. O. Kim, H. O. Kim, H. O. Kim, S. R. Hou, S. R. Hou, S. R. Hou, T. E. Browder, T. E. Browder, N. C. Hastings, H. Kawai, H. Kawai, H. Kawai, P. Kapusta, Hyunwoo Kim, Hyunwoo Kim, Hyunwoo Kim, I. Adachi, I. Adachi, I. Adachi, M. Akatsu, M. Akatsu, M. Akatsu, T. Iijima, S. C. Hsu, S. C. Hsu, Observation of B+ → χc0K+ Physical Review Letters. ,vol. 88, pp. 1- 5 ,(2002)
David Martin Ward Powers, None, Evaluation: from Precision, Recall and F-measure to ROC, Informedness, Markedness and Correlation arXiv: Learning. ,vol. 2, pp. 37- 63 ,(2011)
Vladimir A Protopopescu, Arindam Banerjee, Auroop R Ganguly, Amrudin Agovic, Anomaly Detection in Transportation Corridors using Manifold Embedding ,(2007)
Kathryn Hempstalk, Eibe Frank, Ian H. Witten, One-Class Classification by Combining Density and Class Probability Estimation european conference on machine learning. ,vol. 5211, pp. 505- 519 ,(2008) , 10.1007/978-3-540-87479-9_51
Caren Marzban, The ROC Curve and the Area under It as Performance Measures Weather and Forecasting. ,vol. 19, pp. 1106- 1114 ,(2004) , 10.1175/825.1
Elton P. Hsu, Stochastic analysis on manifolds ,(2002)
Alexander Grigor'yan, Spectral Theory and Geometry: Estimates of heat kernels on Riemannian manifolds Cambridge University Press. pp. 140- 225 ,(1999) , 10.1017/CBO9780511566165.008
Amit Singer, Ronald R. Coifman, Non-linear independent component analysis with diffusion maps Applied and Computational Harmonic Analysis. ,vol. 25, pp. 226- 239 ,(2008) , 10.1016/J.ACHA.2007.11.001
J. A. Hartigan, M. A. Wong, A K-Means Clustering Algorithm Journal of The Royal Statistical Society Series C-applied Statistics. ,vol. 28, pp. 100- 108 ,(1979) , 10.2307/2346830
Hao Huang, Hong Qin, Shinjae Yoo, Dantong Yu, A New Anomaly Detection Algorithm Based on Quantum Mechanics 2012 IEEE 12th International Conference on Data Mining. pp. 900- 905 ,(2012) , 10.1109/ICDM.2012.127