Abstract
Reject option is a technique used to improve classifier’s reliability in decision support systems. It consists in withholding the automatic classification of an item, if the decision is considered not sufficiently reliable. The rejected item is then handled by a different classifier or by a human expert. The vast majority of the works on this issue has been concerned with the development of reject option mechanisms to be used by supervised learning architectures (e.g., MLP, LVQ or SVM). In this paper, however, we aim at proposing alternatives to this view, which are based on the self-organizing map (SOM), originally an unsupervised learning scheme, but that has also been successfully used in the design of prototype-based classifiers. The basic hypothesis we defend is that it is possible to design SOM-based classifiers endowed with reject option mechanisms whose performances are comparable to or better than those achieved by standard supervised classifiers. For this purpose, we carried out a comprehensively evaluation of the proposed SOM-based classifiers on two synthetic and three real-world datasets. The obtained results suggest that the proposed SOM-based classifiers consistently outperform standard supervised classifiers.
Similar content being viewed by others
Notes
There are reject option strategies that are executed during the training of the classifier. These are known as embedded reject option mechanisms. See Sect. 2 for more detail.
At the limit, a classifier with a very low \({\upomega }_r\) would classify only the so-called “easy patterns”.
Available for download at http://www.cis.hut.fi/somtoolbox/.
Values of \({\upomega }_r\) higher than \(0.5\) are equivalent to random guesses.
References
Alhoniemi E, Himberg J, Vesanto J (1999) Probabilistic measures for responses of self-organizing map units. In: Proceedings of the international ICSC congress on computational intelligence methods and applications (CIMA’99). ICSC Academic Press, pp 286–290
Bartlett PL, Wegkamp MH (2008) Classification with a reject option using a Hinge loss. J Mach Learn Res 9:1823–1840
Bellazzi R, Abu-Hanna A (2009) Artificial intelligence in medicine AIME’07. Artif Intell Med 46(1):1–3
Berglund E, Sitte J (2006) Parameterless self-organizing map algorithm. IEEE Trans Neural Netw 17(2):305–316
Biehl M, Ghosh A, Hammer B (2007) Dynamics and generalization ability of LVQ algorithms. J Mach Learn Res 8:323–360
Bounsiar A, Beauseroy P, Grall-Maës E (2008) General solution and learning method for binary classification with performance constraints. Pattern Recognit Lett 29(10):1455–1465
Cardoso JS, Cardoso MJ (2007) Towards an intelligent medical system for the aesthetic evaluation of breast cancer conservative treatment. Artif Intell Med 40:115–126
Cardoso JS, da Costa JFP (2007) Learning to classify ordinal data: the data replication method. J Mach Learn Res 8:1393–1429
Carpenter GA, Grossberg S, Markuzon N, Reynolds JH, Rosen DB (1992) Fuzzy ARTMAP: a neural network architecture for incremental supervised learning of analog multidimensional maps. IEEE Trans Neural Netw 3(5):698–713
Caruana R, Lawrence S, Giles CL (2000) Overfitting in neural nets: backpropagation, conjugate gradient, and early stopping. In: Proceedings of the 2000 neural information processing systems conference (NIPS’00), pp 402–408
Chow C (1970) On optimum recognition error and reject tradeoff. IEEE Trans Inf Theory 16(1):41–46
Cordella L, De Stefano C, Sansone C, Vento M (1995) An adaptive reject option for LVQ classifiers. In: Image analysis and processing, vol. LNCS 974/1995. Springer, pp 68–73
Cordella L, De Stefano C, Tortorella F, Vento M (1995) A method for improving classification reliability of multilayer perceptrons. IEEE Trans Neural Netw 6(5):1140–1147
de Bodt E, Cottrell M, Letremy P, Verleysen M (2004) On the use of self-organizing maps to accelerate vector quantization. Neurocomputing 56:187–203
De Stefano C, Sansone C, Vento M (2000) To reject or not to reject: that is the question—an answer in case of neural classifiers. IEEE Trans Syst Man Cybern C Appl Rev 30(1):574–585
El-Yaniv R, Wiener Y (2010) On the foundations of noise-free selective classification. J Mach Learn Res 11:1605–1641
Flexer A (2001) On the use of self-organizing maps for clustering and visualization. Intell Data Anal 5(5):373–384
Fritzke B (1995) A growing neural gas network learns topologies. In: Tesauro G, Touretzky DS, Leen TK (eds) Advances in neural information processing systems vol 7. MIT Press, Cambridge, pp 625–632
Fu Y, Zhu X, Li B (2013) A survey on instance selection for active learning. Knowl Inf Syst 35(2):249–283
Fumera G, Pillai I, Roli F (2003) Classification with reject option in text categorisation systems. In: Proceedings of the 12th international conference on image analysis and processing (ICIAP’2003). IEEE Computer Society, pp 582–587
Fumera G, Roli F (2002) Support vector machines with embedded reject option. In: Proceedings of the 1st international workshop on pattern recognition with support vector machines (SVM’2002). Springer, pp 68–82
Gama J, de Carvalho AC (2012) Machine learning. In: Machine learning: concepts, methodologies, tools and applications. IGI-Global, pp 13–22. doi:10.4018/978-1-60960-818-7. http://www.igi-global.com/book/machine-learning-concepts-methodologies-tools/50312
Eduardo Gasca A, Sergio Saldaña T, José S. Sánchez G, Valentín Velásquez G, Eréndira Rendón L, Itzel M. Abundez B, Rosa M. Valdovinos R, Rafael Cruz R (2011) A rejection option for the multilayer perceptron using hyperplanes. In: Proceedings of the 10th international conference on adaptive and natural computing algorithms (ICANNGA’2011), vol. LNCS 6593/2011. Springer, pp 51–60
Geebelen D, Suykens J, Vandewalle J (2012) Reducing the number of support vectors of SVM classifiers using the smoothed separable case approximation. IEEE Trans Neural Netw Learn Syst 23(4):682–688
Giles D (2004) Calculating a standard error for the gini coefficient: some further results. Oxford Bull Econ Stat 66(3):124–126
Gini C (1921) Measurement of inequality of incomes. Econ J 31(121):124–126
Goldszmidt M, Cohen I, Fox A, Zhang S (2005) Three research challenges at the intersection of machine learning, statistical induction, and systems. In: Proceedings of the 10th conference on Hot Topics in Operating Systems (HOTOS’05), vol 10, pp 1–6
Guillen A, Herrera LJ, Rubio G, Pomares H, Lendasse A, Rojas I (2010) New method for instance or prototype selection using mutual information in time series prediction. Neurocomputing 73(10–12):2030–2038
Han J, Gao J (2009) Research challenges for data mining in science and engineering. In: Kargupta H, Han J, Yu PS, Motwani R, Kumar V (eds) Next generation of data mining. Chapman & Hall, London, pp 1–18
Hasenjäger M, Ritter H (1998) Active learning with local models. Neural Process Lett 7(2):107–117
Herbei R, Wegkamp MH (2006) Classification with reject option. Can J Stat 34(4):709–721
Holmström L, Hämäläinen A (1993) The self-organizing reduced kernel density estimator. In: Proceedings of the 1993 IEEE International Conference on Neural Networks (ICNN’93), pp 417–421
Ishibuchi H, Nii M (2000) Neural networks for soft decision making. Fuzzy Sets Syst 34(115):121–140
Kohonen T (1982) Self-organized formation of topologically correct feature maps. Biol Cybern 43(1):59–69
Kohonen T (1988) An introduction to neural computing. Neural Netw 1(1):3–16
Kohonen T (1988) The ’neural’ phonetic typewriter. Computer 21(3):11–22
Kohonen T (1990) The self-organizing map. Proc IEEE 78(9):1464–1480
Kohonen T (2001) Self-organizing maps, 3rd edn. Springer, Berlin, Heidelberg, New York
Kohonen T (2003) Learning vector quantization. In: Arbib MA (ed) The handbook of brain theory, neural networks, 2nd edn. MIT Press, Cambridge, pp 631–635
Lau KW, Yin H, Hubbard S (2006) Kernel self-organising maps for classification. Neurocomputing 69:2033–2040
Lotte F, Mouchère H, Lécuyer A (2008) Pattern rejection strategies for the design of self-paced EEG-based brain–computer interfaces. In: Proceedings of the 19th international conference on pattern recognition (ICPR’2008), pp 1–5
Malone J, McGarry K, Wermter S, Bowerman C (2005) Data mining using rule extraction from Kohonen self-organising maps. Neural Comput Appl 15:9–17
Mattos CLC, Barreto GA (2013) ARTIE and MUSCLE models: building ensemble classifiers from fuzzy ART and SOM networks. Neural Comput Appl 22(1):49–61
Oliveira HP, Magalhaes A, Cardoso MJ, Cardoso JS (2010) An accurate and interpretable model for BCCT.core. In: Proceedings of the 32nd annual international conference of the IEEE engineering in medicine and biology Society, pp 6158–6161
Pedreira CE (2006) Learning vector quantization with training data selection. IEEE Trans Pattern Anal Mach Intell 28(1):157–162
Peng H, Zhu S (2007) Handling of incomplete data sets using ICA and SOM in data mining. Neural Comput Appl 16(2):167–172
Ritter H (1991) Asymptotic level density for a class of vector quantization processes. IEEE Trans Neural Netw 2(1):173–175
Riveiro M, Johansson F, Falkman G, Ziemke T (2008) Supporting maritime situation awareness using self organizing maps and gaussian mixture models. In: Proceedings of the 2008 conference on 10th scandinavian conference on artificial intelligence (SCAI’08). IOS Press, pp 84–91
Rocha-Neto AR, Sousa R, Cardoso JS, Barreto GA (2011) Diagnostic of pathology on the vertebral column with embedded reject option. In: Proceedings of the 5th Iberian conference on pattern recognition and image analysis (IbPRIA’2011), vol. LNCS-6669, pp 588–595
Santos-Pereira CM, Pires AM (2005) On optimal reject rules and ROC curves. Pattern Recogn Lett 26(7):943–952
Schleif FM, Villmann T, Hammer B, Schneider P (2011) Efficient kernelized prototype based classification. Int J Neural Syst 21(6):443–57
Seo S, Obermayer K (2002) Soft learning vector quantization. Neural Comput 15:1589–1604
Sim SF, Sági-Kiss V (2011) Multiple self-organising maps (mSOMs) for simultaneous classification and prediction: Illustrated by spoilage in apples using volatile organic profiles. Chemometr Intell Lab Syst 109(1):57–64
Sousa R, Mora B, Cardoso JS (2009) An ordinal data method for the classification with reject option. In: Proceedings of the international conference on machine learning and applications (ICMLA’09), pp 746–750
Sousa R, Rocha Neto AR, Barreto GA, Cardoso JS, Coimbra MT (2014) Reject option paradigm for the reduction of support vectors. In: Proceedings of the 22th European symposium on artificial neural networks, computational intelligence and machine learning (ESANN’2014), pp 1–6
Souza Júnior AH, Barreto GA, Varela AT (2011) A speech recognition system for embedded applications using the SOM and TS-SOM networks. In: Mwasiagi JI (ed) Self-organizing maps—applications and novel algorithm design. InTech Open, Rijeka, pp 97–108. doi:10.5772/14401
Suutala J, Pirttikangas S, Riekki J, Röning J (2004) Reject-optional LVQ-based two-level classifier to improve reliability in footstep identification. In: Ferscha A, Mattern F (eds) Pervasive computing. Springer, Berlin, Heidelberg, pp 182–187
Thomas LC, Edelman DB, Crook JN (2002) Credit scoring and its applications, 1st edn. SIAM, Philadelphia
Tortorella F (2005) A ROC-based reject rule for dichotomizers. Pattern Recognit Lett 26(2):167–180
Turky AM, Ahmad MS (2010) The use of SOM for fingerprint classification. In: IEEE international conference on information retrieval and knowledge management (CAMP’2010), pp 287–290
Umer MF, Khiyal MSH (2007) Classification of textual documents using learning vector quantization. Inf Technol J 6:154–159
Utsugi A (1998) Density estimation by mixture models with smoothing priors. Neural Comput 10:2115–2135
van Hulle M (2012) Self-organizing maps. In: Rozenberg G, Baeck T, Kok J (eds) Handbook of natural computing: theory, experiments, and applications. Springer, Berlin, Heidelberg, pp 585–622
Vasconcelos GC, Fairhurst MC, Bisset DL (1993) Enhanced reliability of multilayer perceptron networks through controlled pattern rejection. Electron Lett 29(3):261–263
Vasconcelos GC, Fairhurst MC, Bisset DL (1995) Investigating feedforward neural networks with respect to the rejection of spurious patterns. Pattern Recognit Lett 16(2):207–212
Villmann T, Haase S (2011) Divergence-based vector quantization. Neural Comput 23(5):1343–1392
Yin H (2008) The self-organizing maps: background, theories, extensions and applications. In: Fulcher J, Jain LC (eds) Computational intelligence: a compendium, studies in computational intelligence, vol 115. Springer, Berlin, Heidelberg, pp 715–762
Yin H, Allinson NM (2001) Self-organizing mixture networks for probability density estimation. IEEE Trans Neural Netw 12(2):405–411
Zidelmal Z, Amirou A, Belouchrani A (2012) Heartbeat classification using support vector machines (SVMs) with an embedded reject option. Int J Pattern Recognit Artif Intell 26(1):1250,001-1–1250,001-17
Acknowledgments
This work was partially supported through Program CNPq/Universidade do Porto/590008/2009-9 and conducted when Ricardo Sousa was in internship at Universidade Federal do Ceará (UFC), Brazil. This work was also partially funded by Fundação para a Ciência e a Tecnologia (FCT)—Portugal through project PTDC/SAU-ENB/114951/2009 and by FEDER funds through the Programa Operacional Factores de Competitividade—COMPETE in the framework of the project PEst-C/SAU/LA0002/2013. The authors also thank Fundação Núcleo de Tecnologia Industrial do Ceará (NUTEC) for providing the laboratorial infrastructure for the execution of the research activities reported in this paper.
Author information
Authors and Affiliations
Corresponding author
Additional information
Ricardo Gamelas Sousa and Ajalmar R. Rocha Neto have contributed equally to this manuscript.
Rights and permissions
About this article
Cite this article
Gamelas Sousa, R., Rocha Neto, A.R., Cardoso, J.S. et al. Robust classification with reject option using the self-organizing map. Neural Comput & Applic 26, 1603–1619 (2015). https://doi.org/10.1007/s00521-015-1822-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-015-1822-2