Landscape analysis for hyperheuristic Bayesian Network structure learning on unseen problems

Bayesian network (BN) structure learning is an NP hard problem. Search and score algorithms are one of the main approaches proposed for learning BN structure from data. Previous research has shown that the relative performances of such algorithms are problem dependent and that fitness landscape analysis can be used to characterize the difficulty of the search for different scoring functions. In this paper, we construct a classifier based on fitness landscape analysis and receiver operating characteristic curves. The classifier labels search landscapes with the most suitable scoring function. We train the classifier on a number of standard benchmark functions. The classifier forms the basis for a selective hyperheuristic algorithm. This uses an initial landscape analysis stage to select a scoring function using the classifier. The hyperheuristic algorithm is tested on a distribution of unseen problems based on mutations of the standard benchmarks. Our results establish that the hyperheuristic performs better than a uniformly random scoring function selection approach that omit the landscape analysis stage. Therefore the effects on performance of problem-dependency can be significantly reduced.

[1]  Matthias Fuchs,et al.  High Performance ATP Systems by Combining Several AI Methods , 1997, IJCAI.

[2]  Guoliang Xue,et al.  Applying two-level simulated annealing on Bayesian structure learning to infer genetic networks , 2004 .

[3]  Bernd Freisleben,et al.  Fitness landscape analysis and memetic algorithms for the quadratic assignment problem , 2000, IEEE Trans. Evol. Comput..

[4]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[5]  James P. Egan,et al.  Signal detection theory and ROC analysis , 1975 .

[6]  David W. Corne,et al.  Evaluating optimization algorithms: bounds on the performance of optimizers on unseen problems , 2011, GECCO '11.

[7]  John A. W. McCall,et al.  Comparative Analysis of Search and Score Metaheuristics for Bayesian Network Structure Learning Using Node Juxtaposition Distributions , 2010, PPSN.

[8]  Jaswinder Pal Singh,et al.  Parallel Implementations of Probabilistic Inference , 1996, Computer.

[9]  Constantin F. Aliferis,et al.  The max-min hill-climbing Bayesian network structure learning algorithm , 2006, Machine Learning.

[10]  John A. W. McCall,et al.  A chain-model genetic algorithm for Bayesian network structure learning , 2007, GECCO '07.

[11]  Wray L. Buntine.  Theory Refinement on Bayesian Networks , 1991, UAI.

[12]  Kent A. Spackman,et al.  Signal Detection Theory: Valuable Tools for Evaluating Inductive Learning , 1989, ML.

[13]  José A. Gámez,et al.  Learning Bayesian networks by Ant Colony Optimisation: searching in two different spaces , 2002 .

[14]  John A. W. McCall,et al.  Two novel Ant Colony Optimization approaches for Bayesian network structure learning , 2010, IEEE Congress on Evolutionary Computation.

[15]  J. Hanley,et al.  A method of comparing the areas under receiver operating characteristic curves derived from the same cases. , 1983, Radiology.

[16]  Brent Boerlage.  Link Strength in Bayesian Networks , 1994 .

[17]  Tao Du,et al.  Efficient Learning Bayesian Networks Using PSO , 2005, CIS.

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

[19]  Lukás Sekanina,et al.  Fitness Landscape Analysis and Image Filter Evolution Using Functional-Level CGP , 2007, EuroGP.

[20]  Qiang Shen,et al.  Learning Bayesian Network Equivalence Classes with Ant Colony Optimization , 2009, J. Artif. Intell. Res..

[21]  G. Uludag,et al.  Fitness landscape analysis of differential evolution algorithms , 2009, 2009 Fifth International Conference on Soft Computing, Computing with Words and Perceptions in System Analysis, Decision and Control.

[22]  Peter A. Flach.  The Geometry of ROC Space: Understanding Machine Learning Metrics through ROC Isometrics , 2003, ICML.

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

[24]  David J. Spiegelhalter,et al.  Local computations with probabilities on graphical structures and their application to expert systems , 1990 .

[25]  John A. W. McCall,et al.  Fitness landscape analysis of Bayesian network structure learning , 2011, 2011 IEEE Congress of Evolutionary Computation (CEC).

[26]  Geoffrey I. Webb,et al.  On the Application of ROC Analysis to Predict Classification Performance Under Varying Class Distributions , 2005, Machine Learning.

[27]  Ricardo del Corazón Grau-Ábalo,et al.  Building Fine Bayesian Networks Aided by PSO-Based Feature Selection , 2007, MICAI.

[28]  J A Swets,et al.  Better decisions through science. , 2000, Scientific American.

[29]  Pedro Larrañaga,et al.  Learning Bayesian network structures by searching for the best ordering with genetic algorithms , 1996, IEEE Trans. Syst. Man Cybern. Part A.

[30]  Pedro Larrañaga,et al.  Structure Learning of Bayesian Networks by Hybrid Genetic Algorithms , 1995, AISTATS.

[31]  Thomas A. Runkler,et al.  Learning of Bayesian networks by a local discovery ant colony algorithm , 2008, 2008 IEEE Congress on Evolutionary Computation (IEEE World Congress on Computational Intelligence).