Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
184 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Inherent Inconsistencies of Feature Importance (2206.08204v2)

Published 16 Jun 2022 in cs.LG, cs.AI, and cs.HC

Abstract: The rapid advancement and widespread adoption of machine learning-driven technologies have underscored the practical and ethical need for creating interpretable artificial intelligence systems. Feature importance, a method that assigns scores to the contribution of individual features on prediction outcomes, seeks to bridge this gap as a tool for enhancing human comprehension of these systems. Feature importance serves as an explanation of predictions in diverse contexts, whether by providing a global interpretation of a phenomenon across the entire dataset or by offering a localized explanation for the outcome of a specific data point. Furthermore, feature importance is being used both for explaining models and for identifying plausible causal relations in the data, independently from the model. However, it is worth noting that these various contexts have traditionally been explored in isolation, with limited theoretical foundations. This paper presents an axiomatic framework designed to establish coherent relationships among the different contexts of feature importance scores. Notably, our work unveils a surprising conclusion: when we combine the proposed properties with those previously outlined in the literature, we demonstrate the existence of an inconsistency. This inconsistency highlights that certain essential properties of feature importance scores cannot coexist harmoniously within a single framework.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (31)
  1. Understanding global feature contributions with additive importance measures. Advances in Neural Information Processing Systems, 33, 2020.
  2. VA Casagrande and IT Diamond. Ablation study of the superior colliculus in the tree shrew (tupaia glis). Journal of Comparative Neurology, 156(2):207–237, 1974.
  3. Understanding the value of features for coreference resolution. In Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing, pages 294–303, Honolulu, Hawaii, October 2008. Association for Computational Linguistics. URL https://www.aclweb.org/anthology/D08-1031.
  4. Rainbow: Combining improvements in deep reinforcement learning. In Thirty-Second AAAI Conference on Artificial Intelligence, 2018.
  5. A unified approach to interpreting model predictions. In Advances in neural information processing systems, pages 4765–4774, 2017.
  6. Marginal contribution feature importance-an axiomatic approach for explaining data. In International Conference on Machine Learning, pages 1324–1335. PMLR, 2021.
  7. Principles of explanatory debugging to personalize interactive machine learning. In Proceedings of the 20th international conference on intelligent user interfaces, pages 126–137, 2015.
  8. Learning global additive explanations for neural nets using model distillation. arXiv preprint arXiv:1801.08640, 2018.
  9. Estimation of feature importance for food intake detection based on random forests classification. In 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pages 6756–6759. IEEE, 2013.
  10. A deep learning approach for cancer detection and relevant gene identification. In Pacific symposium on biocomputing, pages 219–229. World Scientific, 2017.
  11. Machine learning for detecting gene-gene interactions. Applied bioinformatics, 5(2):77–88, 2006.
  12. Bivariate genome-wide association study of depressive symptoms with type 2 diabetes and quantitative glycemic traits. Psychosomatic medicine, 80(3):242, 2018.
  13. Data debugging with shapley importance over end-to-end machine learning pipelines. arXiv preprint arXiv:2204.11131, 2022.
  14. Lemna: Explaining deep learning based security applications. In Proceedings of the 2018 ACM SIGSAC Conference on Computer and Communications Security, pages 364–379, 2018.
  15. Judea Pearl. Causality. Cambridge university press, 2009.
  16. Causal inference, 2010.
  17. From local explanations to global understanding with explainable ai for trees. Nature machine intelligence, 2(1):2522–5839, 2020.
  18. True to the model or true to the data? arXiv preprint arXiv:2006.16234, 2020.
  19. The feature importance ranking measure. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pages 694–709. Springer, 2009.
  20. " why should i trust you?" explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pages 1135–1144, 2016.
  21. Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE international conference on computer vision, pages 618–626, 2017.
  22. On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation. PloS one, 10(7):e0130140, 2015.
  23. Ultra marginal feature importance. arXiv preprint arXiv:2204.09938, 2022.
  24. Jon Kleinberg. An impossibility theorem for clustering. Advances in neural information processing systems, 15, 2002.
  25. Why are we using black box models in ai when we don’t need to? a lesson from an explainable ai competition. 2019.
  26. How interpretable and trustworthy are gams? In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 95–105, 2021.
  27. Statistical stability indices for lime: Obtaining reliable explanations for machine learning models. Journal of the Operational Research Society, 73(1):91–101, 2022.
  28. Zachary C Lipton. The mythos of model interpretability. Queue, 16(3):31–57, 2018.
  29. Problems with shapley-value-based explanations as feature importance measures. In International Conference on Machine Learning, pages 5491–5500. PMLR, 2020.
  30. Eric J Friedman. Paths and consistency in additive cost sharing. International Journal of Game Theory, 32(4):501–518, 2004.
  31. Axiomatic attribution for deep networks. In International conference on machine learning, pages 3319–3328. PMLR, 2017.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now