Learning the Parameters of Possibilistic Networks from Data: Empirical Comparison

Possibilistic networks are belief graphical models based on possibility theory. A possibilistic network either represents experts' epistemic uncertainty or models uncertain information from poor, scarce or imprecise data. Learning possibilistic networks from data in general and from imperfect or scarce datasets in particular, has not received enough attention. This work focuses on parameter learning of possibilistic networks. The main contributions of the paper are i) a study of an extension of the information affinity measure to assess the similarity of possibilistic networks and ii) a comparative empirical evaluation of two approaches for learning the parameters of a possibilistic network from empirical data.

[1]  Christian Borgelt,et al.  Learning possibilistic graphical models from data , 2003, IEEE Trans. Fuzzy Syst..

[2]  Salem Benferhat,et al.  Inference in possibilistic network classifiers under uncertain observations , 2012, Annals of Mathematics and Artificial Intelligence.

[3]  Didier Dubois,et al.  Probability-Possibility Transformations, Triangular Fuzzy Sets, and Probabilistic Inequalities , 2004, Reliab. Comput..

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

[5]  Christian Borgelt,et al.  Learning Probabilistic and Possibilistic Networks : Theory and Applications , 1995 .

[6]  Khaled Mellouli,et al.  Information Affinity: A New Similarity Measure for Possibilistic Uncertain Information , 2007, ECSQARU.

[7]  Mathieu Serrurier,et al.  Learning Structure of Bayesian Networks by Using Possibilistic Upper Entropy , 2014, SMPS.

[8]  Cliff Joslyn,et al.  MEASUREMENT OF POSSIBILISTIC HISTOGRAMS FROM INTERVAL DATA , 1997 .

[9]  Glenn Shafer,et al.  A Mathematical Theory of Evidence , 2020, A Mathematical Theory of Evidence.

[10]  Didier Dubois,et al.  Transforming Probability Intervals into Other Uncertainty Models , 2007, EUSFLAT Conf..

[11]  Didier Dubois,et al.  Practical Methods for Constructing Possibility Distributions , 2016, Int. J. Intell. Syst..

[12]  Salem Benferhat,et al.  Probability-Possibility Transformations: Application to Credal Networks , 2015, SUM.

[13]  Fabio Gagliardi Cozman,et al.  Credal networks , 2000, Artif. Intell..

[14]  Thierry Denoeux,et al.  Inferring a possibility distribution from empirical data , 2006, Fuzzy Sets Syst..

[15]  Christian Borgelt,et al.  Graphical Models: Representations for Learning, Reasoning and Data Mining , 2009 .

[16]  Nahla Ben Amor,et al.  Learning possibilistic networks from data: a survey , 2015, IFSA-EUSFLAT.

[17]  D. Dubois,et al.  On Possibility/Probability Transformations , 1993 .

[18]  Nahla Ben Amor,et al.  Probability-Possibility Transformation: - Application to Bayesian and Possibilistic Networks , 2013, WILF.

[19]  Salem Benferhat,et al.  On the Analysis of Probability-Possibility Transformations: Changing Operations and Graphical Models , 2015, ECSQARU.

[20]  Didier Dubois,et al.  Possibility Theory - An Approach to Computerized Processing of Uncertainty , 1988 .

[21]  Marco Zaffalon,et al.  Learning Reliable Classifiers From Small or Incomplete Data Sets: The Naive Credal Classifier 2 , 2008, J. Mach. Learn. Res..