Classification by ordinal sums of conjunctive and disjunctive functions for explainable AI and interpretable machine learning solutions

作者：

Highlights：

• Classification by neural networks is highly effective, but the solution is not re-traceable, hence interpretable.

• Fuzzy classification is re-traceable, but a higher number of sets reduces interpretability.

• A novel classification by the ordinal sums of conjunctive and disjunctive functions is proposed.

摘要

•Classification by neural networks is highly effective, but the solution is not re-traceable, hence interpretable.•Fuzzy classification is re-traceable, but a higher number of sets reduces interpretability.•A novel classification by the ordinal sums of conjunctive and disjunctive functions is proposed.

论文关键词：Explainable AI,Interpretable Machine Learning (ML),Interactive ML,Aggregation functions,Ordinal sums,Glass-box,Transparency

论文评审过程：Received 19 August 2020, Revised 24 January 2021, Accepted 26 February 2021, Available online 2 March 2021, Version of Record 9 March 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.106916