• Corpus ID: 11338479

Convex Hull-Based Multi-objective Genetic Programming for Maximizing ROC Performance

@article{Wang2013ConvexHM,
  title={Convex Hull-Based Multi-objective Genetic Programming for Maximizing ROC Performance},
  author={Pu Wang and M. Emmerich and Rui Li and Ke Tang and Thomas B{\"a}ck and Xin Yao},
  journal={ArXiv},
  year={2013},
  volume={abs/1303.3145}
}
ROC is usually used to analyze the performance of classifiers in data mining. ROC convex hull (ROCCH) is the least convex major-ant (LCM) of the empirical ROC curve, and covers potential optima for the given set of classifiers. Generally, ROC performance maximization could be considered to maximize the ROCCH, which also means to maximize the true positive rate (tpr) and minimize the false positive rate (fpr) for each classifier in the ROC space. However, tpr and fpr are conflicting with each… 
2 Citations
A multi-level knee point based multi-objective evolutionary algorithm for AUC maximization
TLDR
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-objectives optimization.
Co-evolutionary multi-population genetic programming for classification in software defect prediction: An empirical case study
TLDR
This paper uses colonization and migration operators along with three ensemble selection strategies for the multi-objective evolutionary algorithm and compares the performance of the operators for software defect prediction datasets with varying levels of data imbalance.

References

SHOWING 1-10 OF 45 REFERENCES
Multiobjective genetic programming for maximizing ROC performance
TLDR
This paper proposes multiobjective genetic programming (MOGP) to obtain a group of nondominated classifiers, with which the maximum ROCCH can be achieved, and proposes a memetic approach into GP by defining two local search strategies specifically designed for classification problems.
A new multi-objective evolutionary algorithm based on convex hull for binary classifier optimization
TLDR
A novel population- based multi-objective evolutionary algorithm (MOEA) for binary classifier optimization and how the Pareto front approximation generated by the proposed MOEA is better than the one generated by NSGA-II, one of the most known and used population-based MOEAs.
Multi-objective genetic fuzzy classifiers for imbalanced and cost-sensitive datasets
TLDR
The FRBC selected from the convex hull produced by the three-objective optimization approach achieves the lowest classification cost among the techniques used as comparison in two specific medical applications.
Convex hull ranking algorithm for multi-objective evolutionary algorithms
TLDR
This paper uses convex hull concepts to present a new ranking procedure for multi-objective evolutionary algorithms, and applies it as an alternative ranking procedure to NSGA-II for non-dominated comparisons, and test it using some benchmark problems.
A multi-objective genetic programming approach to developing Pareto optimal decision trees
  • H. Zhao
  • Computer Science
    Decis. Support Syst.
  • 2007
TLDR
This paper proposes a multi-objective genetic programming approach to developing such alternative Pareto optimal decision trees and allows the decision maker to specify partial preferences on the conflicting objectives to further reduce the number of alternative solutions.
The Multi-objective Differential Evolution Algorithm Based on Quick Convex Hull Algorithms
The convex hull of a set of points is the smallest convex set that contains the points. This article presents a multi-objective differential evolutionary algorithm based on quick convex hull
A Memetic Genetic Programming with decision tree-based local search for classification problems
In this work, we propose a new genetic programming algorithm with local search strategies, named Memetic Genetic Programming(MGP), for classification problems. MGP aims to acquire a classifier with
Robust Classification for Imprecise Environments
TLDR
It is shown that it is possible to build a hybrid classifier that will perform at least as well as the best available classifier for any target conditions, and in some cases, the performance of the hybrid actually can surpass that of the best known classifier.
A novel diversification strategy for multi-objective evolutionary algorithms
TLDR
A new archiving strategy based on the Convex Hull of Individual Minima (CHIM), which is intended to maintain a well-distributed set of non-dominated solutions is introduced.
Multi-Objective Genetic Programming for Classification with Unbalanced Data
TLDR
This paper proposes a Multi-Objective Genetic Programming (MOGP) approach to evolve a Pareto front of classifiers along the optimal trade-off surface representing minority and majority class accuracy for binary class imbalance problems and shows that a diverse set of solutions was found.
...
1
2
3
4
5
...