Possibilistic Network-Based Classifiers: On the Reject Option and Concept Drift Issues

In this paper, we deal with two important issues regarding possibilistic network-based classifiers. The first issue addresses the reject option in possibilistic network-based classifiers. We first focus on simple threshold-based reject rules and provide interpretations for the ambiguity and distance reject then introduce a third reject kind named incompleteness reject occurring when the inputs are missing or incomplete. The second important issue we address is the one of concept drift. More specifically, we propose an efficient solution for revising a possibilistic network classifier with new information.

[1]  Jeff Z. Pan,et al.  An Argument-Based Approach to Using Multiple Ontologies , 2009, SUM.

[2]  C. K. Chow,et al.  On optimum recognition error and reject tradeoff , 1970, IEEE Trans. Inf. Theory.

[3]  Eyke Hüllermeier,et al.  Flexible Control of Case-Based Prediction in the Framework of Possibility Theory , 2000, EWCBR.

[4]  Alexey Tsymbal,et al.  The problem of concept drift: definitions and related work , 2004 .

[5]  Rudolf Kruse,et al.  Knwoledge revision in Markov networks , 2004 .

[6]  Gerhard Widmer,et al.  Learning in the Presence of Concept Drift and Hidden Contexts , 1996, Machine Learning.

[7]  Hoel Le Capitaine,et al.  An Optimum Class-Rejective Decision Rule and Its Evaluation , 2010, 2010 20th International Conference on Pattern Recognition.

[8]  Thien M. Ha,et al.  The Optimum Class-Selective Rejection Rule , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[9]  Khaled Mellouli,et al.  A Two-Steps Algorithm for Min-Based Possibilistic Causal Networks , 2001, ECSQARU.

[10]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[11]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems - networks of plausible inference , 1991, Morgan Kaufmann series in representation and reasoning.

[12]  Michael Clarke,et al.  Symbolic and Quantitative Approaches to Reasoning and Uncertainty , 1991, Lecture Notes in Computer Science.

[13]  Nir Friedman,et al.  Bayesian Network Classifiers , 1997, Machine Learning.

[14]  R. Jeffrey.  The Logic of Decision , 1984 .

[15]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[16]  Salem Benferhat,et al.  An Efficient Algorithm for Naive Possibilistic Classifiers with Uncertain Inputs , 2008, SUM.

[17]  Khaled Mellouli,et al.  Naïve possibilistic network classifiers , 2009, Fuzzy Sets Syst..

[18]  Didier Dubois,et al.  A synthetic view of belief revision with uncertain inputs in the framework of possibility theory , 1997, Int. J. Approx. Reason..

[19]  Hoel Le Capitaine,et al.  Classification with Reject Options in a Logical Framework: a fuzzy residual implication approach , 2009, IFSA/EUSFLAT Conf..

[20]  Carl Frélicot.  On Unifying Probabilistic/Fuzzy and Possibilistic Rejection-Based Classifiers , 1998, SSPR/SPR.

[21]  Christian Borgelt,et al.  Graphical models - methods for data analysis and mining , 2002 .

[22]  R. Casey,et al.  Advances in Pattern Recognition , 1971 .

[23]  E. Hisdal.  Conditional possibilities independence and noninteraction , 1978 .

[24]  Philippe Leray,et al.  Pertinence des mesures de confiance en classification , 2000 .