Output details
11 - Computer Science and Informatics
Coventry University
FSVM-CIL: Fuzzy support vector machines for class imbalance learning
<24> The paper presents a method of improving SVMs for class imbalance learning (called Fuzzy-SVM-CIL), to successfully handle the class imbalance in the presence of outliers and noise. It was the first paper, when published, addressing the problem of class imbalance for FSVMs. The proposed FSVM-CIL method has been thoroughly validated on several real-world imbalanced datasets. The technique presented in the paper has influenced and is a component part in the development of many recent methods on class imbalance for SVMs, as evident from citations. The work has been funded by a prestigious Clarendon PhD scholarship award at Oxford University.