Using Model Explanations to Guide Deep Learning Models Towards Consistent Explanations for EHR Data

Watson, M; Awwad Shekh Hasan, B; Al Moubayed, N

Abstract

It has been shown that identical Deep Learning (DL) architectures will produce distinct explanations when trained with different hyperparameters that are orthogonal to the task (e.g. random seed, training set order). In domains such as healthcare and finance, where transparency and explainability is paramount, this can be a significant barrier to DL adoption. In this study we present a further analysis of explanation (in)consistency on 6 tabular datasets/tasks, with a focus on Electronic Health Records data. We propose a novel deep learning ensemble architecture that trains its sub-models to produce consistent explanations, improving explanation consistency by as much as 315% (e.g. from 0.02433 to 0.1011 on MIMIC-IV), and on average by 124% (e.g. from 0.12282 to 0.4450 on the BCW dataset). We evaluate the effectiveness of our proposed technique and discuss the implications our results have for both industrial applications of DL and explainability as well as future methodological work.

Citation

Watson, M., Awwad Shekh Hasan, B., & Al Moubayed, N. (2022). Using Model Explanations to Guide Deep Learning Models Towards Consistent Explanations for EHR Data. Scientific Reports, 12(19899), Article 19899. https://doi.org/10.1038/s41598-022-24356-6

• Journal Article Type: Article
• Acceptance Date: Nov 14, 2022
• Online Publication Date: Nov 18, 2022
• Publication Date: 2022
• Deposit Date: Nov 17, 2022
• Publicly Available Date: Feb 8, 2023
• Journal: Scientific Reports
• Electronic ISSN: 2045-2322
• Publisher: Nature Research
• Peer Reviewed: Peer Reviewed
• Volume: 12
• Issue: 19899
• Article Number: 19899
• DOI: https://doi.org/10.1038/s41598-022-24356-6
• Public URL: https://durham-repository.worktribe.com/output/1185319