作者: Jie Lu , Guangquan Zhang , Ning Lu
DOI:
关键词:
摘要: Competence enhancement plays an important role in case-base editing. Traditional competence methods tend to omit the evolving nature of a case-based learner, but take whole as static training set. This may seriously delay or even prohibit learner from learning new concepts, when concept drifts. paper proposes Modified Blame Based Noise Removal algorithm (M-BBNR). Our MBBNR preserves some potential noise cases, case representing novel concepts. Experiment show that with such "wait-and-see" policy, developed M-BBNR outperforms other famous on real world dataset and is able tuning according drift effectively.