Abstract
Recently, numerous high-performance machine learning models have been proposed. Unfortunately, such models often produce black-box decisions derived using opaque reasons and logic. Therefore, it is important to develop a tool that automatically gives the reasons underlying the black-box model’s decision. Ideally, the tool should be model-agnostic: applicable to any machine-learning model without knowing model details. A well-known previous work, LIME, is based on the linear decision. Although LIME provides important features for the decision, the result is still difficult to understand for users because the result might not contain the features required for the decision. We propose a novel model-agnostic explanation method named MP-LIME. The explanation consists of feature sets, each of which can reconstruct the decision correctly. Thereby, users can easily understand each feature set. By comparing our method to LIME, we demonstrate that our method often improves precision drastically. We also provide practical examples in which our method provides reasons for the decisions.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Baehrens, D., Schroeter, T., Harmeling, S., Kawanabe, M., Hansen, K., Müller, K.R.: How to explain individual classification decisions. J. Mach. Learn. Res. 11, 1803–1831 (2010). http://dl.acm.org/citation.cfm?id=1756006.1859912
Batal, I., Hauskrecht, M.: Constructing classification features using minimal predictive patterns. In: Proceedings of the 19th ACM International Conference on Information and Knowledge Management, pp. 869–878. ACM (2010)
Van den Bergh, M., Boix, X., Roig, G., Van Gool, L.: Seeds: Superpixels extracted via energy-driven sampling. Int. J. Comput. Vision 111(3), 298–314 (2015)
Blitzer, J., Dredze, M., Pereira, F.: Biographies, bollywood, boom-boxes and blenders: domain adaptation for sentiment classification. In: Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics, pp. 440–447 (2007)
Caruana, R., Lou, Y., Gehrke, J., Koch, P., Sturm, M., Elhadad, N.: Intelligible models for healthcare: Predicting pneumonia risk and hospital 30-day readmission. In: Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1721–1730. ACM (2015)
Efron, B., Hastie, T., Johnstone, I., Tibshirani, R., et al.: Least angle regression. Ann. Stat. 32(2), 407–499 (2004)
Guidotti, R., et al.: A survey of methods for explaining black box models. ACM Comput. Surv. (CSUR) 51(5), 93 (2018)
Hara, S., Ishihata, M.: Approximate and exact enumeration of rule models. In: AAAI (2018)
Hara, S., Maehara, T.: Enumerate lasso solutions for feature selection. In: AAAI, pp. 1985–1991 (2017)
Lakkaraju, H., et al.: A machine learning framework to identify students at risk of adverse academic outcomes. In: Proceedings of the 21st ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1909–1918. ACM (2015)
Lakkaraju, H., Bach, S.H., Leskovec, J.: Interpretable decision sets: a joint framework for description and prediction. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1675–1684. ACM (2016)
Ribeiro, M.T., Singh, S., Guestrin, C.: 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, pp. 1135–1144. ACM (2016)
Ribeiro, M.T., Singh, S., Guestrin, C.: Anchors: high-precision model-agnostic explanations. In: AAAI Conference on Artificial Intelligence (2018)
Ruggieri, S.: Enumerating distinct decision trees. In: International Conference on Machine Learning, pp. 2960–2968 (2017)
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., Wojna, Z.: Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2818–2826 (2016)
Tibshirani, R.: Regression shrinkage and selection via the lasso. J. Roy. Stat. Soc. Ser. B (Methodological) 58, 267–288 (1996)
Wang, F., Rudin, C.: Falling rule lists. In: Artificial Intelligence and Statistics, pp. 1013–1022 (2015)
Acknowledgments
I would like to thank Quentin Labernia Louis Marie and Nguyen Van Quang for their comments on the manuscript. This work was partially supported by JSPS Kakenhi 15H02665 and 17K00002.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Asano, K., Chun, J., Koike, A., Tokuyama, T. (2019). Model-Agnostic Explanations for Decisions Using Minimal Patterns. In: Tetko, I., Kůrková, V., Karpov, P., Theis, F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Theoretical Neural Computation. ICANN 2019. Lecture Notes in Computer Science(), vol 11727. Springer, Cham. https://doi.org/10.1007/978-3-030-30487-4_19
Download citation
DOI: https://doi.org/10.1007/978-3-030-30487-4_19
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-30486-7
Online ISBN: 978-3-030-30487-4
eBook Packages: Computer ScienceComputer Science (R0)