Knowledge discovery from noisy imbalanced and incomplete binary class data

作者：

Highlights：

• Analysis of 84 different models in noisy imbalanced and incomplete datasets.

• MICE and KNN imputation techniques outperform in incomplete imbalanced datasets.

• A result shows SMOTE-ENN better in noisy imbalanced datasets.

• MICE-SMOTE-ENN performs better in noisy imbalanced and incomplete datasets.

摘要

•Analysis of 84 different models in noisy imbalanced and incomplete datasets.•MICE and KNN imputation techniques outperform in incomplete imbalanced datasets.•A result shows SMOTE-ENN better in noisy imbalanced datasets.•MICE-SMOTE-ENN performs better in noisy imbalanced and incomplete datasets.

论文关键词：Missing value imputation techniques,Oversampling techniques,Noise,Binary class imbalanced data,Performance metrics

论文评审过程：Received 23 March 2020, Revised 11 March 2021, Accepted 7 May 2021, Available online 15 May 2021, Version of Record 20 May 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115179