Abstract
Noise detection algorithms commonly face the problems of over-cleansing, large computational complexity and long response time. Preserving the original structure of data is critical for any classifier, whereas over-cleansing will adversely affect the quality of data. Besides, the high time complexity remains one of the main defects for most noise detectors, especially those exhibiting an ensemble structure. Moreover, numerous studies reported that ensemble-based techniques outperform other techniques in the accuracy of noisy instances identification. In the present study, a fast class noise detector called FMF (fast class noise detector with multi-factor-based learning) was proposed. FMF, three existing ensemble-based noise detectors and the case with original data were compared on 10 classification tasks. C5.0 acted as the classifiers to access the efficiency of these algorithms. As revealed from the results, the FMF shortened the processing time at least twelve times by three baselines, and it achieved the highest accuracy in most cases.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Garcia, L., de Carvalho, A., Lorena, A.C.: Effect of label noise in the complexity of classification problems. Neurocomputing 160, 108–119 (2015)
Hu, Z., Li, B., Hu, Y.: Fast sign recognition with weighted hybrid k-nearest neighbors based on holistic features from local feature descriptors. J. Comput. Civ. Eng. 31(5) (2017). https://doi.org/10.1061/(ASCE)CP.1943-5487.0000673
Liu, F.T., Ting, K.M., Zhou, Z.H.: Isolation forest. In: Proceeding of 2008 Eighth IEEE International Conference on Data Mining, pp. 413–422 (2008) Location (1999). https://doi.org/10.1109/ICDM.2008.17
Sáez, J., Galar, M., Luengo, J., Herrera, F.: Analyzing the presence of noise in multi-class problems: alleviating its influence with the one vs one decomposition. Knowl. Inf. Syst. 38, 1–28 (2014)
Chen, J., Zhang, C., Xue, X., Liu, C.L.: Fast instance selection for speeding up support vector machines. Knowl. Based Syst. 45, 1–7 (2013)
Rathee, S., Ratnoo, S., Ahuja, J.: Instance selection using multi-objective chc evolutionary algorithm. In: Fong, S., Akashe, S., Mahalle, P.N. (eds.) Information and Communication Technology for Competitive Strategies. LNNS, vol. 40, pp. 475–484. Springer, Singapore (2019). https://doi.org/10.1007/978-981-13-0586-3_48
de Haro-GarcÃa, A., Pérez-RodrÃguez, J., GarcÃa-Pedrajas, N.: Combining three strategies for evolutionary instance selection for instance-based learning. Swarm Evol. Comput. 42, 160–172 (2018)
Sáez, J.A., Galar, M., Luengo, J., Herrera, F.: INFFC: an iterative class noise filter based on the fusion of classifiers with noise sensitivity control. Inf. Fusion 27, 19–32 (2016)
Garcia, L., Lehmann, J., Carvalho, A., Lorena, A.: New label noise injection methods for the evaluation of noise filters. Knowl. Based Syst. 163, 693–704 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Zheng, W., Jin, M. (2020). A Fast Class Noise Detector with Multi-factor-based Learning. In: Chellappan, S., Choo, KK.R., Phan, N. (eds) Computational Data and Social Networks. CSoNet 2020. Lecture Notes in Computer Science(), vol 12575. Springer, Cham. https://doi.org/10.1007/978-3-030-66046-8_2
Download citation
DOI: https://doi.org/10.1007/978-3-030-66046-8_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-66045-1
Online ISBN: 978-3-030-66046-8
eBook Packages: Computer ScienceComputer Science (R0)