Variable Ordering for Bayesian Networks Learning from Data

Classification is an important task in data mining processes. In this work, the χ test is used to define the order of the variables of a dataset to be used in Bayesian classification tasks. Two Bayesian classifiers are used to verify the influence of the variables ordering in the classification rate. The first one is based on the K2 algorithm which has strong dependency upon the initial order of the variables, and the second one is the algorithm used in BNPC software which is based on the conditional independence (CI) principle and doesn’t depend on an initial variables order. Four datasets (from UCI repository) are classified with and without the defined order and the results are compared.

[1]  Catherine Blake,et al.  UCI Repository of machine learning databases , 1998 .

[2]  P. Spirtes,et al.  Causation, prediction, and search , 1993 .

[3]  Judea Pearl,et al.  Equivalence and Synthesis of Causal Models , 1990, UAI.

[4]  David Maxwell Chickering,et al.  Optimal Structure Identification With Greedy Search , 2002, J. Mach. Learn. Res..

[5]  Wai Lam,et al.  LEARNING BAYESIAN BELIEF NETWORKS: AN APPROACH BASED ON THE MDL PRINCIPLE , 1994, Comput. Intell..

[6]  Weiru Liu,et al.  An algorithm for Bayesian network construction from data , 1997 .

[7]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems , 1988 .

[8]  Luis M. de Campos,et al.  A new approach for learning belief networks using independence criteria , 2000, Int. J. Approx. Reason..

[9]  Moninder Singh,et al.  Construction of Bayesian network structures from data: A brief survey and an efficient algorithm , 1995, Int. J. Approx. Reason..

[10]  Joe Suzuki,et al.  Learning Bayesian Belief Networks Based on the MDL Principle : An Efficient Algorithm Using the Branch and Bound Technique , 1999 .

[11]  Estevam R. Hruschka,et al.  Missing values prediction with K2 , 2002, Intell. Data Anal..

[12]  David Maxwell Chickering,et al.  Learning Bayesian Networks: The Combination of Knowledge and Statistical Data , 1994, Machine Learning.

[13]  Marek J. Druzdzel,et al.  A Hybrid Anytime Algorithm for the Construction of Causal Models From Sparse Data , 1999, UAI.

[14]  Luis M. de Campos,et al.  Independency relationships and learning algorithms for singly connected networks , 1998, J. Exp. Theor. Artif. Intell..

[15]  N. Wermuth,et al.  Graphical and recursive models for contingency tables , 1983 .

[16]  Alice M. Agogino,et al.  Automated Construction of Sparse Bayesian Networks from Unstructured Probabilistic Models and Domain Information , 2013, UAI.

[17]  P. Spirtes,et al.  An Algorithm for Fast Recovery of Sparse Causal Graphs , 1991 .

[18]  David J. Spiegelhalter,et al.  Sequential updating of conditional probabilities on directed graphical structures , 1990, Networks.

[19]  Gregory F. Cooper,et al.  A Bayesian Method for the Induction of Probabilistic Networks from Data , 1992 .

[20]  Hiroshi Motoda,et al.  Feature Selection for Knowledge Discovery and Data Mining , 1998, The Springer International Series in Engineering and Computer Science.

[21]  Luis M. de Campos,et al.  A hybrid methodology for learning belief networks: BENEDICT , 2001, Int. J. Approx. Reason..

[22]  Estevam R. Hruschka,et al.  Ordering attributes for missing values prediction and data classification , 2002 .