Towards Adapting XCS for Imbalance Problems

Nguyen, T.H.; Foitong, Sombut; Srinil, Phaitoon; Pinngern, Ouen

doi:10.1007/978-3-540-89197-0_102

Cited by 6 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Baseline XCSR performed very poorly in highly output-biased environments. By tuning XCSR and also disabling either types of subsumption, performance was improved and acceptable and comparable to two other XCSR variants designed based on the suggestions of Nguyen et al [4] and Orriols-Puig and Bernadó-Mansilla [5]. However, all of these three variants only managed to get acceptable performances provided that they are trained for a very large number of trials, specially for the highly output-biased problems (for instance, about 2.2 million trials needed at the 96% output bias value).…”

Section: Output Biasmentioning

confidence: 75%

On the problems of using learning classifier systems for fraud detection

Behdad

French

Barone

et al. 2010

Proceedings of the 12th Annual Conference on Genetic and Evolutionary Computation

View full text Add to dashboard Cite

Fraud detection problems have some uniquely challenging properties which make them difficult. In this paper, we investigate the fraud detection problem by describing the common properties of electronic fraud and examining how learning classifier systems (LCSs) can be applied to it. Also, we introduce "random Boolean function" (RBF); an abstract problem with high level of controllability which can be tuned to exhibit those characteristics individually, and report the results of using XCSR (a continuous variant of LCS) on RBF problem and also on a real-world problem. Results from our experiments demonstrate that XCSR can overcome most of the difficulties inherent to the fraud detection problem and can achieve good performance in case of the real-world problem.

show abstract

Section: Output Biasmentioning

confidence: 75%

On the problems of using learning classifier systems for fraud detection

Behdad

French

Barone

et al. 2010

Proceedings of the 12th Annual Conference on Genetic and Evolutionary Computation

View full text Add to dashboard Cite

show abstract

“…3 and 4 are two variants of XCSR specifically modified to handle output-biased environments. The first, by Nguyen et al [8], modifies three baseline XCSR control parameters: two of which are set to different constant values (α: 0.05, ν: 10) and the third ( 0) is calculated based on the inverse of the output bias level. Its performance is plotted using a brown solid line decorated with circles.…”

Section: Experiments 3: Output Biasmentioning

confidence: 99%

An investigation of real-valued accuracy-based learning classifier systems for electronic fraud detection

Behdad

Barone

French

et al. 2010

Proceedings of the 12th Annual Conference Companion on Genetic and Evolutionary Computation

View full text Add to dashboard Cite

Fraud is a serious problem that costs the worldwide economy billions of dollars annually. However, fraud detection is difficult as perpetrators actively attempt to masquerade their actions, among typically overwhelming large volumes of, legitimate activity. In this paper, we investigate the fraud detection problem and examine how learning classifier systems can be applied to it. We describe the common properties of fraud, introducing an abstract problem which can be tuned to exhibit those characteristics. We report experiments on this abstract problem with a popular real-time learning classifier system algorithm; results from our experiments demonstrating that this approach can overcome the difficulties inherent to the fraud detection problem. Finally we apply the algorithm to a real-world problem and show that it can achieve good performance in this domain.

show abstract