Generative adversarial minority enlargement - A local linear over-sampling synthetic method

Ke Wang, Tongqing Zhou, Menghua Luo, Xionglve Li, Zhiping Cai

Abstract: Highlights•A local linear property is proposed.•Logistic regression is integrated into a generative and adversarial process.•Enlarging generating data space by the mixture of majority and minority samples.•Better performances than 12 baselines by 4 classifiers on 28 datasets.