Improving the undersampling technique by optimizing the termination condition for software defect prediction

Shuo Feng, Jacky Keung, Yan Xiao, Peichang Zhang, Xiao Yu, Xiaochun Cao

Abstract: Highlights•Proposing a novel learning-to-rank undersampling technique.•Further improving the undersampling technique by refining the termination condition.•Proving the current way to terminate the undersampling technique is improper.•Revealing the trade-off lying in the undersampling technique.