A multi-level knee point based multi-objective evolutionary algorithm for AUC maximization

The area under receiver operating characteristic curve (AUC) is one of the widely used metrics for measuring imbalanced data classification results. Designing multi-objective evolutionary algorithms for AUC maximization problem has attracted much attention of researchers recently. However, most of these methods either search the Pareto front directly, or perform tailored convex hull search for AUC maximization. None of them take the advantage of multi-level knee points found in the process of evolution for AUC maximization. To this end, this paper proposes a multi-level knee point based multi-objective evolutionary algorithm (named MKnEA-AUC) for AUC maximization on the basis of a recently developed knee point driven evolutionary algorithm for multi/many-objective optimization. In MKnEA-AUC, an adaptive clustering strategy is proposed for automatically determining the knee points on the current population. By utilizing the preference of found knee points, the evolution of the population can converge quickly. We verify the effectiveness of the proposed algorithm MKnEA-AUC on 13 widely used benchmark data sets and the experimental results demonstrate that MKnEA-AUC is superior over the state-of-the-art algorithms for AUC maximization.

[1]  George W. Quinn,et al.  Modest proposals for improving biometric recognition papers , 2015, 2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS).

[2]  Kalyanmoy Deb,et al.  Multi-objective optimization using evolutionary algorithms , 2001, Wiley-Interscience series in systems and optimization.

[3]  Ye Tian,et al.  An Efficient Approach to Nondominated Sorting for Evolutionary Multiobjective Optimization , 2015, IEEE Transactions on Evolutionary Computation.

[4]  Xin Yao,et al.  Convex Hull-Based Multiobjective Genetic Programming for Maximizing Receiver Operating Characteristic Performance , 2015, IEEE Transactions on Evolutionary Computation.

[5]  Mark Johnston,et al.  Evolving Diverse Ensembles Using Genetic Programming for Classification With Unbalanced Data , 2013, IEEE Transactions on Evolutionary Computation.

[6]  Ye Tian,et al.  A Decision Variable Clustering-Based Evolutionary Algorithm for Large-Scale Many-Objective Optimization , 2018, IEEE Transactions on Evolutionary Computation.

[7]  Yves Lecourtier,et al.  A multi-model selection framework for unknown and/or evolutive misclassification cost problems , 2010, Pattern Recognit..

[8]  Ye Tian,et al.  PlatEMO: A MATLAB Platform for Evolutionary Multi-Objective Optimization [Educational Forum] , 2017, IEEE Computational Intelligence Magazine.

[9]  Qingfu Zhang,et al.  MOEA/D: A Multiobjective Evolutionary Algorithm Based on Decomposition , 2007, IEEE Transactions on Evolutionary Computation.

[10]  Thorsten Joachims,et al.  A support vector method for multivariate performance measures , 2005, ICML.

[11]  Bernhard Sendhoff,et al.  A Multiobjective Evolutionary Algorithm Using Gaussian Process-Based Inverse Modeling , 2015, IEEE Transactions on Evolutionary Computation.

[12]  Jinhua Zheng,et al.  Spread Assessment for Evolutionary Multi-Objective Optimization , 2009, EMO.

[13]  Dezheng Zhang,et al.  Optimizing area under the ROC curve via extreme learning machines , 2017, Knowl. Based Syst..

[14]  Ye Tian,et al.  A Knee Point-Driven Evolutionary Algorithm for Many-Objective Optimization , 2015, IEEE Transactions on Evolutionary Computation.

[15]  Xin Yao,et al.  Convex Hull-Based Multi-objective Genetic Programming for Maximizing ROC Performance , 2013, ArXiv.

[16]  Sean J. Welleck,et al.  Efficient AUC Optimization for Information Ranking Applications , 2015, ECIR.

[17]  R. Lyndon While,et al.  A faster algorithm for calculating hypervolume , 2006, IEEE Transactions on Evolutionary Computation.

[18]  Kalyanmoy Deb,et al.  A combined genetic adaptive search (GeneAS) for engineering design , 1996 .

[19]  Beatrice Lazzerini,et al.  A new multi-objective evolutionary algorithm based on convex hull for binary classifier optimization , 2007, 2007 IEEE Congress on Evolutionary Computation.

[20]  James E. Goin,et al.  ROC curve estimation and hypothesis testing: applications to breast cancer detection , 1982, Pattern Recognit..

[21]  Bernhard Sendhoff,et al.  Generalization Improvement in Multi-Objective Learning , 2006, The 2006 IEEE International Joint Conference on Neural Network Proceedings.

[22]  Tom Fawcett,et al.  Robust Classification for Imprecise Environments , 2000, Machine Learning.

[23]  Kalyanmoy Deb,et al.  A fast and elitist multiobjective genetic algorithm: NSGA-II , 2002, IEEE Trans. Evol. Comput..

[24]  Xin Yao,et al.  Multiobjective genetic programming for maximizing ROC performance , 2014, Neurocomputing.

[25]  Ke Tang,et al.  Convex hull-based multi-objective evolutionary computation for maximizing receiver operating characteristics performance , 2015, Memetic Computing.

[26]  Beatrice Lazzerini,et al.  Multi-objective genetic fuzzy classifiers for imbalanced and cost-sensitive datasets , 2010, Soft Comput..

[27]  Matthew A. Kupinski,et al.  Multiobjective Genetic Optimization of Diagnostic Classifiers with Implications for Generating ROC Curves , 1999, IEEE Trans. Medical Imaging.

[28]  Ye Tian,et al.  An Indicator-Based Multiobjective Evolutionary Algorithm With Reference Point Adaptation for Better Versatility , 2018, IEEE Transactions on Evolutionary Computation.

[29]  Kalyanmoy Deb,et al.  Simulated Binary Crossover for Continuous Search Space , 1995, Complex Syst..

[30]  Jianchao Zeng,et al.  A fitness approximation assisted competitive swarm optimizer for large scale expensive optimization problems , 2018, Memetic Comput..