The von Mises Naive Bayes Classifier for Angular Data

Directional and angular information are to be found in almost every field of science. Directional statistics provides the theoretical background and the techniques for processing such data, which cannot be properly managed by classical statistics. The von Mises distribution is the best known angular distribution. We extend the naive Bayes classifier to the case where directional predictive variables are modeled using von Mises distributions. We find the decision surfaces induced by the classifiers and illustrate their behavior with artificial examples. Two applications to real data are included to show the potential uses of these models. Comparisons with classical techniques yield promising results.

[1]  W. L. Kovach,et al.  Quantitative methods for the study of lycopod megaspore ultrastructure , 1989 .

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

[3]  Marvin Minsky,et al.  Steps toward Artificial Intelligence , 1995, Proceedings of the IRE.

[4]  Nir Friedman,et al.  Probabilistic Graphical Models - Principles and Techniques , 2009 .

[5]  Pedro M. Domingos,et al.  Beyond Independence: Conditions for the Optimality of the Simple Bayesian Classifier , 1996, ICML.

[6]  Mehran Sahami,et al.  Learning Limited Dependence Bayesian Classifiers , 1996, KDD.

[7]  K. Mardia,et al.  Protein Bioinformatics and Mixtures of Bivariate von Mises Distributions for Angular Data , 2007, Biometrics.

[8]  Peter B. Krenesky,et al.  Protein Geometry Database: a flexible engine to explore backbone conformations and their relationships to covalent geometry , 2009, Nucleic Acids Res..

[9]  Pedro M. Domingos,et al.  On the Optimality of the Simple Bayesian Classifier under Zero-One Loss , 1997, Machine Learning.

[10]  Philipp Berens,et al.  CircStat: AMATLABToolbox for Circular Statistics , 2009, Journal of Statistical Software.

[11]  Jesper Ferkinghoff-Borg,et al.  A generative, probabilistic model of local protein structure , 2008, Proceedings of the National Academy of Sciences.

[12]  Olivier Pourret,et al.  Bayesian networks : a practical guide to applications , 2008 .

[13]  Michael J. Pazzani,et al.  Searching for Dependencies in Bayesian Classifiers , 1995, AISTATS.

[14]  Usama M. Fayyad,et al.  Multi-Interval Discretization of Continuous-Valued Attributes for Classification Learning , 1993, IJCAI.

[15]  Ronald E. Goldstein,et al.  Principles and techniques , 2009 .

[16]  Nicholas I. Fisher,et al.  Statistical Analysis of Circular Data , 1993 .

[17]  S. R. Jammalamadaka,et al.  Directional Statistics, I , 2011 .

[18]  Remco R. Bouckaert,et al.  Estimating replicability of classifier learning experiments , 2004, ICML.

[19]  Nir Friedman,et al.  Probabilistic Graphical Models , 2009, Data-Driven Computational Neuroscience.

[20]  Mark A. Peot,et al.  Geometric Implications of the Naive Bayes Assumption , 1996, UAI.

[21]  Kanti V. Mardia,et al.  Bayesian analysis for bivariate von Mises distributions , 2010 .

[22]  F. Perrin,et al.  Étude mathématique du mouvement brownien de rotation , 1928 .

[23]  Richard O. Duda,et al.  Pattern classification and scene analysis , 1974, A Wiley-Interscience publication.