Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2412.18952 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909441197080576 |
|---|---|
| author | Nayyem, Navid Rakin, Abdullah Wang, Longwei |
| author_facet | Nayyem, Navid Rakin, Abdullah Wang, Longwei |
| contents | This paper explores the intricate relationship between interpretability and robustness in deep learning models. Despite their remarkable performance across various tasks, deep learning models often exhibit critical vulnerabilities, including susceptibility to adversarial attacks, over-reliance on spurious correlations, and a lack of transparency in their decision-making processes. To address these limitations, we propose a novel framework that leverages Local Interpretable Model-Agnostic Explanations (LIME) to systematically enhance model robustness. By identifying and mitigating the influence of irrelevant or misleading features, our approach iteratively refines the model, penalizing reliance on these features during training. Empirical evaluations on multiple benchmark datasets demonstrate that LIME-guided refinement not only improves interpretability but also significantly enhances resistance to adversarial perturbations and generalization to out-of-distribution data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_18952 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Bridging Interpretability and Robustness Using LIME-Guided Model Refinement Nayyem, Navid Rakin, Abdullah Wang, Longwei Machine Learning Artificial Intelligence This paper explores the intricate relationship between interpretability and robustness in deep learning models. Despite their remarkable performance across various tasks, deep learning models often exhibit critical vulnerabilities, including susceptibility to adversarial attacks, over-reliance on spurious correlations, and a lack of transparency in their decision-making processes. To address these limitations, we propose a novel framework that leverages Local Interpretable Model-Agnostic Explanations (LIME) to systematically enhance model robustness. By identifying and mitigating the influence of irrelevant or misleading features, our approach iteratively refines the model, penalizing reliance on these features during training. Empirical evaluations on multiple benchmark datasets demonstrate that LIME-guided refinement not only improves interpretability but also significantly enhances resistance to adversarial perturbations and generalization to out-of-distribution data. |
| title | Bridging Interpretability and Robustness Using LIME-Guided Model Refinement |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2412.18952 |