Optimizing the early glaucoma detection from visual fields by combining preprocessing techniques and ensemble classifier with selection strategies

作者：

Highlights：

• Machine learning is a major issue for automatic diagnosis of glaucoma.

• We have fused preprocessing and bagging ensemble with selection strategy.

• Approach and tests are made according to a three-level fusion.

• The full-fusion has been most effective for early glaucoma.

• Weak learner SVM and SMOTE have improved performances.

摘要

•Machine learning is a major issue for automatic diagnosis of glaucoma.•We have fused preprocessing and bagging ensemble with selection strategy.•Approach and tests are made according to a three-level fusion.•The full-fusion has been most effective for early glaucoma.•Weak learner SVM and SMOTE have improved performances.

论文关键词：Ensemble classifier,Bagging,Static selection,Dynamic selection,Features selection,Visual fields,Glaucoma,Unbalanced data

论文评审过程：Received 27 April 2020, Revised 24 July 2021, Accepted 23 September 2021, Available online 4 October 2021, Version of Record 27 October 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115975