Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content
Springer Nature Link
Log in

Partially supervised Independent Factor Analysis using soft labels elicited from multiple experts: application to railway track circuit diagnosis

  • Focus
  • Published:
Soft Computing Aims and scope Submit manuscript

Abstract

Using a statistical model in a diagnosis task generally requires a large amount of labeled data. When ground truth information is not available, too expensive or difficult to collect, one has to rely on expert knowledge. In this paper, it is proposed to use partial information from domain experts expressed as belief functions. Expert opinions are combined in this framework and used with measurement data to estimate the parameters of a statistical model using a variant of the EM algorithm. The particular application investigated here concerns the diagnosis of railway track circuits. A noiseless Independent Factor Analysis model is postulated, assuming the observed variables extracted from railway track inspection signals to be generated by a linear mixture of independent latent variables linked to the system component states. Usually, learning with this statistical model is performed in an unsupervised way using unlabeled examples only. In this paper, it is proposed to handle this learning process in a soft-supervised way using imperfect information on the system component states. Fusing partially reliable information about cluster membership is shown to significantly improve classification results.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8

Similar content being viewed by others

Explore related subjects

Discover the latest articles, news and stories from top researchers in related subjects.

References

  • Amari S, Cichocki A, Yang HH (1996) A new learning algorithm for blind signal separation. In: Proceedings of the 8th conference on advances in neural information processing systems (NIPS). MIT Press, Cambridge, pp 756–763

  • Ambroise C, Denoeux T, Govaert G, Smets P (2001) Learning from an imprecise teacher: probabilistic and evidential approaches. In: Proceedings of the 10th international symposium on applied stochastic models and data analysis (ASMDA), Compiègne, France, pp 100–105.

  • Ambroise C, Govaert G (2000) EM algorithm for partially known labels. In: Proceedings of the 7th conference of the international federation of classification societies (IFCS). Springer, Namur, Belgium, pp 161–166

  • Amini R, Gallinari P (2005) Semi-supervised learning with an imperfect supervisor. Knowl Inf Syst 8(4):385–413

    Article  Google Scholar 

  • Attias H (1999) Independent factor analysis. Neural Comput 11(4):803–851

    Article  Google Scholar 

  • Bartholomew DJ, Martin K (1999) Latent variable models and factor analysis. 2nd edn. Arnold, London

    MATH  Google Scholar 

  • Bell AJ, Sejnowski TJ (1995) An information maximization approach to blind separation and blind deconvolution. Neural Comput 7(6):1129–1159

    Article  Google Scholar 

  • Ben Yaghlane A, Denoeux T, Mellouli K (2006) Elicitation of expert opinions for constructing belief functions. In: Proceedings of the 11th international conference on information processing and management of uncertainty in knowledge-based systems (IPMU ’06), Paris, France, pp 403–411

  • Bishop CM (2006) Pattern recognition and machine learning (information science and statistics). Springer, New York

    Google Scholar 

  • Coelho F, de Pádua Braga A, Natowicz R, Rouzier R (2010) Semi-supervised model applied to the prediction of the response to preoperative chemotherapy for breast cancer. Soft Comput A Fusion Found Methodol Appl 15(6):1137–1144

  • Chapelle O, Scholkopf B, Zien A (2006) Semi-supervised learning. MIT Press, Cambridge

    Google Scholar 

  • Côme E, Cherfi Z, Oukhellou L, Aknin P (2008) Semi-supervised IFA with prior knowledge on the mixing process: an application to a railway device diagnosis. In: Proceedings of the 7th ICMLA’08. San Diego, pp 415–420

  • Côme E, Oukhellou L, Denoeux T, Aknin P (2009a) Noiseless Independent Factor Analysis with mixing constraints in a semi-supervised framework. Application to railway device fault diagnosis. In: Proceedings of the 19th international conference on artificial neural networks (ICANN), Limassol, Cyprus, pp 416–425

  • Côme E, Oukhellou L, Denoeux T, Aknin P (2009b) Learning from partially supervised data using mixture models and belief functions. Pattern Recognit 42(3):334–348

    Article  MATH  Google Scholar 

  • Côme E, Oukhellou L, Denoeux T, Aknin P (2011) Fault diagnosis of a railway device using semi-supervised independent factor analysis with mixing constraints. Pattern Anal Appl (to appear). doi:10.1007/s10044-011-0212-3

  • Comon P (1994) Independent component analysis, a new concept? Signal Process 36(3):287–314 (special issue on higher-order statistics)

    Article  MATH  Google Scholar 

  • Dempster AP (1967) Upper and lower probabilities induced by a multivalued mapping. Ann Math Stat 38(2):325–339

    Article  MathSciNet  MATH  Google Scholar 

  • Denoeux T (1995) A k-nearest neighbor classification rule based on Dempster–Shafer theory. IEEE Trans Syst Man Cybern 25(5):804–813

    Article  Google Scholar 

  • Denoeux T (2008) Conjunctive and disjunctive combination of belief functions induced by non distinct bodies of evidence. Artif Intell 172:234–264

    Article  MathSciNet  MATH  Google Scholar 

  • Denoeux T (2010) Maximum likelihood from evidential data: an extension of the EM algorithm. In: Borgelt C et al (eds) Combining soft computing and statistical methods in data analysis, AISC 77. Springer, pp 181–188

  • Denoeux T (2011) Maximum likelihood estimation from uncertain data in the belief function framework. IEEE Trans Knowl Data Eng. doi:10.1109/TKDE.2011.201

  • Denoeux T, Zouhal LM (2001) Handling possibilistic labels in pattern classification using evidential reasoning. Fuzzy Sets Syst 122(3):47–62

    Article  MathSciNet  Google Scholar 

  • Dubois D, Prade H (1988) Representation and combination of uncertainty with belief functions and possibility measures. Comput Intell 4(4):244–264

    Article  Google Scholar 

  • Duda RO, Hart PE, Stork DG (2001) Pattern classification. 2nd edn. Wiley, New York

    MATH  Google Scholar 

  • Elouedi Z, Mellouli K, Smets Ph (2004) Assessing sensor reliability for multisensor data fusion within the Transferable Belief Model. IEEE Trans Syst Man Cybern B 34(1):782–787

    Article  Google Scholar 

  • Elouedi Z, Mellouli K, Smets Ph (2001) Belief decision trees: theoretical foundations. Int J Approx Reason 28(2-3):91–124

    Article  MathSciNet  MATH  Google Scholar 

  • Ghahramani Z (2004) Unsupervised learning. In: Bousquet O, Raetsch G, von Luxburg U (eds) Advanced lectures on machine learning. Springer, Berlin, pp 72–112

  • Grandvalet Y (2002) Logistic regression for partial labels. In: Proceedings of the 9th international conference on information processing and management of uncertainty in knowledge-based systems (IPMU), vol 3. Annecy, France, pp 1935–1941

  • Grandvalet Y, Bengio Y (2005) Semi-supervised learning by entropy minimization. In: Proceedings of the 17th conference on advances in neural information processing systems (NIPS). MIT Press, Cambridge, pp 529–536

  • Ha-Duong M (2008) Hierarchical fusion of expert opinions in the Transferable Belief Model, application to climate sensitivity. Int J Approx Reason 49(3):555–574

    Article  Google Scholar 

  • Hastie T, Tibshirani R, Friedman J (2006) The elements of statistical learning, data mining, inference and prediction. Springer, New York

    Google Scholar 

  • Hüllermeier E, Beringer J (2005) Learning from ambiguously labeled examples. In: Proceedings of the 6th international symposium on intelligent data analysis (IDA-05), Madrid, Spain, pp 168–179

  • Jenhani I, Ben Amor N, Elouedi Z (2007) Decision trees as possibilistic classifiers. Int J Approx Reason 43(8):784–807

    Google Scholar 

  • Jraidi I, Elouedi Z (2007) Belief classification approach based on generalized credal EM. In: Mellouli K (ed) 9th European conference on symbolic and quantitative approaches to reasoning with uncertainty (ECSQARU ’07), Springer, Hammamet, Tunisia, pp 524–535

  • Klose A (2004) Extracting fuzzy classification rules from partially labeled data. Soft Comput A Fusion Found Methodol Appl 8(6):417–427

    Article  Google Scholar 

  • Lawrence ND, Schölkopf B(2001) Estimating a kernel fisher discriminant in the presence of label noise. In: Proceedings of the 18th international conference on machine learning (ICML). Morgan Kaufmann, San Francisco, pp 306–313

  • Li Y, Wessels L, De Ridder D, Reinders M (2007) Classification in the presence of class noise using a probabilistic kernel fisher method. Pattern Recognit 40(12):3349–3357

    Article  MATH  Google Scholar 

  • McLachlan GJ (1977) Estimating the linear discriminant function from initial samples containing a small number of unclassified observations. J Am Stat Assoc 72(358):403–406

    MathSciNet  MATH  Google Scholar 

  • McLachlan GJ, Krishnan T (1997) The EM algorithm and extension. Wiley, New York

    Google Scholar 

  • Mercier D, Quost B, Denoeux T (2008) Refined modeling of sensor reliability in the belief function framework using contextual discounting. Inf Fusion 9(2):246–258

    Article  Google Scholar 

  • Moulines E, Cardoso J, Cassiat E (1997) Maximum likelihood for blind separation and deconvolution of noisy signals using mixture models. In: Proceedings of the IEEE international conference on acoustics, speech and signal processing, Munich, Germany, pp 3617–3620

  • Nocedal J, Wright S (1999) Numerical optimization. In: Springer series in operations research and financial engineering. Springer, Berlin

  • Oukhellou L, Debiolles A, Aknin P, Vilette F (2004) Automatic diagnostic of track circuit in a predictive maintenance context. In: International conference on railway engineering, London

  • Oukhellou L, Debiolles A, Denoeux T, Aknin P (2010) Fault diagnosis in railway track circuits using Dempster–Shafer classifier fusion. Eng Appl Artif Intell 23:117–128

    Article  Google Scholar 

  • Palacios A, Sánchez L, Couso I (2011) Linguistic cost-sensitive learning of genetic fuzzy classifiers for imprecise data. Int J Approx Reason 52(6):841–862

    Article  Google Scholar 

  • Pichon F, Denoeux T (2010) The unnormalized Dempster’s rule of combination: a new justification from the Least Commitment Principle and some extensions. J Autom Reason 45(1):61–87

    Article  MathSciNet  MATH  Google Scholar 

  • Quost B, Masson M-H, Denoeux T (2011) Classifier fusion in the Dempster–Shafer framework using optimized t-norm based combination rules. Int J Approx Reason 52(3):353–374

    Article  MathSciNet  Google Scholar 

  • Shafer G (1976) A mathematical theory of evidence. University Press, Princeton

    MATH  Google Scholar 

  • Smets Ph (1990) The combination of evidence in the Transferable Belief Model. IEEE Trans Pattern Anal Mach Intell 12(5):447–458

    Article  Google Scholar 

  • Smets Ph (1993) Belief functions: the disjunctive rule of combination and the generalized Bayesian theorem. Int J Approx Reason 9:1–35

    Article  MathSciNet  MATH  Google Scholar 

  • Smets Ph (1995) The canonical decomposition of a weighted belief. In: International joint conference on artificial intelligence. Morgan Kaufman, San Mateo, CA, pp 1896–1901

  • Smets Ph, Kennes R (1994) The transferable belief model. Artif Intell 66:191–234

    Article  MathSciNet  MATH  Google Scholar 

  • Vannoorenbergue P, Denoeux T (2002) Handling uncertain labels in multiclass problems using belief decision trees. In: Proceedings of IPMU’2002, vol III. Annecy, France, pp 1919–1926

  • Vannoorenberghe P, Smets Ph (2005) Partially supervised learning by a credal EM approach. In: Godo L (ed) Proceedings of the 8th European conference on symbolic and quantitative approaches to reasoning with uncertainty (ECSQARU ’05), Springer, Barcelona, Spain, pp 956–967

  • Worden K, Manson G, Denoeux T (2009) An evidence-based approach to damage location on an aircraft structure. Mech Syst Signal Process 23(6):1792–1804

    Article  Google Scholar 

  • Yager RR (1987) On the Dempster–Shafer framework and new combination rules. Inf Sci 41(2):93–137

    Article  MathSciNet  MATH  Google Scholar 

Download references

Acknowledgments

This work was supported by the French National Research Agency (ANR) under project DIAGHIST. The authors thank the French National Railway Company (SNCF) and its experts for their collaboration.

Author information

Authors and Affiliations

  1. GRETTIA, French Institute of Science and Technology for Transport, Development and Networks, Université Paris-Est, Marne-la-Vallée, France

    Zohra L. Cherfi, Etienne Côme & Patrice Aknin

  2. LISSI, Université Paris-Est Créteil, Créteil, France

    Latifa Oukhellou

  3. HEUDIASYC, Université de Technologie de Compiègne, UMR CNRS 6599, Compiègne, France

    Thierry Denœux

Authors
  1. Zohra L. Cherfi
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Latifa Oukhellou
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Etienne Côme
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Thierry Denœux
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Patrice Aknin
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Zohra L. Cherfi.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Cherfi, Z.L., Oukhellou, L., Côme, E. et al. Partially supervised Independent Factor Analysis using soft labels elicited from multiple experts: application to railway track circuit diagnosis. Soft Comput 16, 741–754 (2012). https://doi.org/10.1007/s00500-011-0766-4

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00500-011-0766-4

Keywords

Search

Navigation