Abstract
In some real applications, it requires to recognize known class(es) as well as unknown class. Traditional classification can only recognize the known class(es) which occur(s) in the training set. Recently, Kernel Null Foley-Sammon Transformation (KNFST) is proposed to overcome the weakness that traditional classification cannot identify unknown class. In KNFST, the samples from the same class are mapped to a point via null projected directions. For a test sample, the predicted label is the class with the minimum distance to it in the projected space. If the test sample is far away from all projected points, it comes from an unknown class. However, KNFST only considers the global information which is depicted by between- and within- class scatter matrices. When the internal structure is complex, KNFST may fail. In order to handle this issue, this paper proposes a Kernel Null Regularized Foley-Sammon Transformation (KNRFST) by introducing a regularized term into KNFST. The regularized term is depicted by a neighborhood which consists of reverse nearest neighbors. The neighborhood can depict the local information in the class. The experimental results, performed on several datasets, show that KNRFST is superior to some previous methods, including KNFST, Local KNFST, WOC-SVM, Deep SVDD and OCC-GAN.
Similar content being viewed by others
Notes
k(xi, xj) = exp (2κHIK(xi, xj) − κHIK(xi, xi) − κHIK(xj, xj)).
These datasets are collected from https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/
References
Chung H, Ko H, Lee H et al (2020) Deep learning for heart rate estimation from reflectance photoplethysmography with acceleration power spectrum and acceleration intensity. IEEE Access 8:63390–63402
Asghari-Esfeden S, Sznaier M, Camps O (2020) Dynamic motion representation for human action recognition. The IEEE Winter Conference on Applications of Computer Vision:557–566
Chalapathy R, Chawla S (2019) Deep learning for anomaly detection: a survey. arXiv preprint arXiv:1901.03407, 2019
Pimentel MAF, Clifton DA, Clifton L et al (2014) A review of novelty detection. Signal Process 99:215–249
Gao X, Sun Q, Xu H et al (2020) Sparse and collaborative representation based kernel pairwise linear regression for image set classification. Expert Syst Appl 120:112886
Liu Z, Yue L, Lai Z, Weihua O, Zhang K (2021) Robust sparse low-rank embedding for image dimension reduction. Appl Soft Comput 113:20211129
Ghosal T, Edithal V, Ekbal A et al (2021) Is your document novel? Let attention guide you. An attention-based model for document-level novelty detection. Nat Lang Eng 27(4):427–454
Adarsh S, Asharaf S, Anoop VS (2022) Sentence-level document novelty detection using latent dirichlet allocation with auto-encoders. International Conference on Innovative Computing and Communications. Springer, Singapore, pp 511–519
Gruhl C, Sick B, Tomforde S (2021) Novelty detection in continuously changing environments. Futur Gener Comput Syst 114:138–154
Silva JCS, Macdo D, Zanchettin C, et al (2021) Multi-class mobile money service financial fraud detection by integrating supervised learning with adversarial autoencoders. Int Joint Conf Neural Networks (IJCNN). IEEE, 2021: 1–7
Kumari R, Ashok N, Ghosal T et al (2021) Misinformation detection using multitask learning with mutual learning for novelty detection and emotion recognition. Inf Process Manag 58(5):102631
Kumari R, Ashok N, Ghosal T et al (2022) What the fake? Probing misinformation detection standing on the shoulder of novelty and emotion. Inf Process Manag 59(1):102740
Maes K, Salens W, Feremans G et al (2022) Anomaly detection in long-term tunnel deformation monitoring. Eng Struct 250:113383
Zheng H, Shi Z, Zhou C, Haardt M, Chen J (2021) Coupled coarray tensor CPD for DOA estimation with coprime L-shaped array. IEEE Signal Process Lett 28:15451549
Yue W, Li C, Mao G et al (2021) Evolution of road traffic congestion control: a survey from perspective of sensing, communication, and computation. China Communications 18(12):151–177
Yue W, Li C, Wang S et al (2022) Towards enhanced recovery and system stability: analytical solutions for dynamic incident effects in road networks. IEEE Trans Intell Transp Syst 23(1):483–798
Chaudhuri NB, Dhar D, Yammiyavar PG (2022) Automating assessment of design exams: a case study of novelty evaluation. Expert Syst Appl 189:116108
Bodesheim P, Freytag A, Rodner E et al (2013) Kernel null space methods for novelty detection. Proc IEEE Conf Comput Vis Pattern Recognit:3374–3381
Zhu F, Gao J, Yang J et al (2022) Neighborhood linear discriminant analysis. Pattern Recogn 123:108422
Amarbayasgalan T, Jargalsaikhan B, Ryu KH (2018) Unsupervised novelty detection using deep autoencoders with density based clustering. Appl Sci 8(9):1468
Daniel T, Kurutach T, Tamar A (2019) Deep Variational semi-supervised novelty detection. arXiv preprint arXiv:1911.04971, 2019
Jumutc V, Suykens JAK, Supervised novelty detection. (2013) IEEE symposium on computational intelligence and data mining (CIDM). IEEE 2013:143–149
Scholkopf B, Platt JC, Shawe-Taylor J, Smola AJ, Williamson RC (2001) Estimating the support of a high-dimensional distribution. Neural Comput 13(7):14431471
Tax DMJ, Duin RPW (2004) Support vector data description. Mach Learn 54(1):4566
Zhu F, Ning Y, Chen X et al (2021) On removing potential redundant constraints for SVOR learning. Appl Soft Comput 102:106941
Pisner DA, Schnyer DM (2020) Support vector machine. Mach Learn Acad Press:101–121
Zhu F, Yang J, Xu S et al (2016) Relative density degree induced boundary detection for one-class SVM. Soft Comput 20(11):4473–4485
Zhu F, Ye N, Wei Y et al (2014) Boundary detection and sample reduction for one-class support vector machines. Neurocomputing 123:166–173
Zhu F, Yang J, Gao C et al (2016) A weighted one-class support vector machine. Neurocomputing 189:1–10
Kemmler M, Rodner E, Wacker E-S, Denzler J (2013) One-class classification with Gaussian processes. Pattern Recogn 46(12):35073518
Qian Leng, Honggang Qi, Jun Miao, Wentao Zhu, and Guiping Su. One-class classification with extreme learning machine. Math Probl Eng, 2015:111, 2015
Ruff L, Vandermeulen R, Goernitz N et al (2018) Deep one-class classification. Int Conf Mach Learn:4393–4402
Liznerski P, Ruff L, Vandermeulen R A, et al (2020) Explainable deep one-class classification. arXiv preprint arXiv:2007.01760, 2020
Yang Y, Hou C, Lang Y et al (2019) One-class classification using generative adversarial networks. IEEE Access 7:37970–37979
Sabokrou M, Khalooei M, Fathy M et al (2018) Adversarially learned one-class classifier for novelty detection. Proc IEEE Conf Comput Vis Pattern Recognit:3379–3388
Landgrebe T, Paclk P, Tax DMJ et al (2005) Optimising two-stage recognition systems. International Workshop on Multiple Classifier Systems. Springer, Berlin, Heidelberg, pp 206–215
Tax DMJ, Duin RPW (2008) Growing a multi-class classifier with a reject option. Pattern Recogn Lett 29(10):1565–1570
Chan FTS, Wang ZX, Patnaik S et al (2020) Ensemble-learning based neural networks for novelty detection in multi-class systems. Appl Soft Comput 93:106396
Silva SR, Vieira T, Martnez D et al (2021) On novelty detection for multi-class classification using non-linear metric learning. Expert Syst Appl 167:114193
Zhang L, Xiang T, Gong S (2016) Learning a discriminative null space for person reidentification. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016: 1239–1248
Liu J, Lian Z, Wang Y et al (2017) Incremental kernel null space discriminant analysis for novelty detection. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE 2017:4123–4131
Huang X, Xu J, Guo G (2018) Incremental kernel null Foley-Sammon transform for person re-identification. 2018 24th International Conference on Pattern Recognition (ICPR). IEEE, pp 1683–1688
Ali TMF (2018) Chaudhuri S. Maximum margin metric learning over discriminative null space for person re-identification. Proceedings of the European Conference on Computer Vision (ECCV). 2018:122–138
Dong H, Lu P, Liu C et al (2018) Person re-identification by kernel null space marginal fisher analysis. Pattern Recogn Lett 107:66–74
Li S, Liu X, Liu W, et al (2016) A discriminative null space based deep learning approach for person re-identification. 2016 4th International Conference on Cloud Computing and Intelligence Systems (CCIS). IEEE, 2016:480–48, Beijing
Guo YF, Wu L, Lu H et al (2006) Null Foley-Sammon transform. Pattern Recogn 39(11):2248–2251
Korn F, Muthukrishnan S (2000) Influence sets based on reverse nearest neighbor queries. In Proceedings of the ACM SIGMOD International Conference on Management of Data, 2000, 29(2):201-212.
Deng C, Yang E, Liu T et al (2019) Unsupervised semantic-preserving adversarial hashing for image search. IEEE Trans Image Process 28(8):4032–4044
Bodesheim P, Freytag A, Rodner E et al (2015) Local novelty detection in multi-class recognition problems. 2015 IEEE Winter Conference on Applications of Computer Vision. IEEE, 2015:813–820
Biagetti G, Crippa P, Falaschetti L et al (2020) Dataset from PPG wireless sensor for activity monitoring. Data Brief 29:105044
Guo Z, Min A, Yang B et al (2021) A sparse oblique-manifold nonnegative matrix factorization for hyperspectral unmixing[J]. IEEE Trans Geosci Remote Sens 60:1–13
Lin X, Wu J, Mumtaz S et al (2020) Blockchain-based on-demand computing resource trading in IoV-assisted smart city. IEEE Trans Emerg Top Comput 9(3):1373–1385
Li J, Zhou Z, Wu J et al (2019) Decentralized on-demand energy supply for blockchain in internet of things: a microgrids approach. IEEE Trans Comput Soc Sys 6(6):1395–1406
Zhang W, Zhang B, Zhu W et al (2021) Comprehensive assessment of MODIS-derived near-surface air temperature using wide elevation-spanned measurements in China. Sci Total Environ 800:149535
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Li, X., Pu, B. Regularized supervised novelty detection and its application in activity monitoring. Appl Intell 53, 4813–4826 (2023). https://doi.org/10.1007/s10489-022-03782-z
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-022-03782-z