×

zbMATH — the first resource for mathematics

AI Explainability 360: an extensible toolkit for understanding data and machine learning models. (English) Zbl 07255161
Summary: As artificial intelligence algorithms make further inroads in high-stakes societal applications, there are increasing calls from multiple stakeholders for these algorithms to explain their outputs. To make matters more challenging, different personas of consumers of explanations have different requirements for explanations. Toward addressing these needs, we introduce AI Explainability 360, an open-source Python toolkit featuring ten diverse and state-of-the-art explainability methods and two evaluation metrics. Equally important, we provide a taxonomy to help entities requiring explanations to navigate the space of interpretation and explanation methods, not only those in the toolkit but also in the broader literature on explainability. For data scientists and other users of the toolkit, we have implemented an extensible software architecture that organizes methods according to their place in the AI modeling pipeline. The toolkit is not only the software, but also guidance material, tutorials, and an interactive web demo to introduce AI explainability to different audiences. Together, our toolkit and taxonomy can help identify gaps where more explainability methods are needed and provide a platform to incorporate them as they are developed.
MSC:
68T05 Learning and adaptive systems in artificial intelligence
PDF BibTeX XML Cite
Full Text: Link
References:
[1] ELI5: Explain like I’m five. GitHub repository,https://github.com/TeamHG-Memex/eli5.
[2] EthicalML-XAI: An explainability toolbox for machine learning. GitHub repository,https:// github.com/EthicalML/xai.
[3] H2O.ai: Machine learning interpretability resources. GitHub repository,https://github.com/ h2oai/mli-resources.
[4] Skater: Python library for model interpretation/explanations.GitHub repository,https:// github.com/oracle/Skater.
[5] tf-explain: Interpretability methods for tf.keras models with Tensorflow 2.0. GitHub repository, https://github.com/sicara/tf-explain.
[6] Maximilian Alber, Sebastian Lapuschkin, Philipp Seegerer, Miriam H¨agele, Gr´egoire Montavon Kristof T. Sch¨utt, Wojciech Samek, Sven D¨ahne Klaus-Robert M¨uller, and Pieter-Jan Kindermans. iNNvestigate neural networks! arXiv:1808.04260, 2018.
[7] David Alvarez-Melis and Tommi Jaakkola.Towards robust interpretability with self-explaining neural networks. InAdvances in Neural Information Processing Systems, pages 7775-7784. 2018.
[8] Vijay Arya, Rachel K. E. Bellamy, Pin-Yu Chen, Amit Dhurandhar, Michael Hind, Samuel C. Hoffman, Stephanie Houde, Q. Vera Liao, Ronny Luss, Aleksandra Mojsilovi´c, Sami Mourad, Pablo Pedemonte, Ramya Raghavendra, John Richards, Prasanna Sattigeri, Karthikeyan Shanmugam, Moninder Singh, Kush R. Varshney, Dennis Wei, and Yunfeng Zhang. One explanation does not fit all: A toolkit and taxonomy of AI explainability techniques. arXiv:1909.03012, 2019.
[9] Hubert Baniecki and Przemyslaw Biecek. modelStudio: Interactive studio with explanations for ML predictive models.Journal of Open Source Software, 4(43):1798, 2019.
[10] Przemys law Biecek. DALEX: Explainers for complex predictive models in R.Journal of Machine Learning Research, 19(84):1-5, 2018.
[11] Sanjeeb Dash, Oktay G¨unl¨uk, and Dennis Wei. Boolean decision rules via column generation. In Advances in Neural Information Processing Systems, pages 4655-4665, 2018.
[12] Amit Dhurandhar, Pin-Yu Chen, Ronny Luss, Chun-Chen Tu, Paishun Ting, Karthikeyan Shanmugam, and Payel Das. Explanations based on the missing: Towards contrastive explanations with pertinent negatives. InAdvances in Neural Information Processing Systems, pages 592-603, 2018a.
[13] Amit Dhurandhar, Karthikeyan Shanmugam, Ronny Luss, and Peder Olsen. Improving simple models with confidence profiles. InAdvances in Neural Information Processing Systems, pages 10296-10306, 2018b.
[14] FICO. FICO Explainable Machine Learning Challenge.https://community.fico.com/s/ explainable-machine-learning-challenge, 2018.
[15] Alex A. Freitas. Comprehensible classification models — a position paper.ACM SIGKDD Explorations, 15(1):1-10, 2014.
[16] Karthik Gurumoorthy, Amit Dhurandhar, Guillermo Cecchi, and Charu Aggarwal. Efficient data representation by selecting prototypes with importance weights.InProceedings of the IEEE International Conference on Data Mining, 2019.
[17] Michael Hind, Dennis Wei, Murray Campbell, Noel C. F. Codella, Amit Dhurandhar, Aleksandra Mojsilovic, Karthikeyan Natesan Ramamurthy, and Kush R. Varshney. TED: Teaching AI to explain its decisions. InProceedings of the AAAI/ACM Conference on AI, Ethics, and Society, pages 123-129, 2019.
[18] Janis Klaise, Arnaud Van Looveren, Giovanni Vacanti, and Alexandru Coca. Alibi: Algorithms for monitoring and explaining machine learning models, February 2020. URLhttps://github. com/SeldonIO/alibi.
[19] Narine Kokhliyan, Edward Wang, Vivek Miglani, and Orion Richardson.Captum.In https://github.com/pytorch/captum, 2019.
[20] Abhishek Kumar, Prasanna Sattigeri, and Avinash Balakrishnan. Variational inference of disentangled latent concepts from unlabeled observations. InProceedings of the International Conference on Learning Representations, 2018.
[21] Scott M. Lundberg and Su-In Lee. A unified approach to interpreting model predictions. InAdvances in Neural Information Processing Systems, pages 4765-4774, 2017.
[22] Ronny Luss, Pin-Yu Chen, Amit Dhurandhar, Prasanna Sattigeri, Karthik Shanmugam, and Chun-Chen Tu.Generating contrastive explanations with monotonic attribute functions. arXiv:1905.12698, 2019.
[23] Christoph Molnar, Giuseppe Casalicchio, and Bernd Bischl. iml: An r package for interpretable machine learning.Journal of Open Source Software, 3(26):786, 2018. doi: 10.21105/joss.00786. URLhttps://doi.org/10.21105/joss.00786.
[24] Harsha Nori, Samuel Jenkins, Paul Koch, and Rich Caruana. InterpretML: A unified framework for machine learning interpretability. arXiv:1909.09223, 2019.
[25] Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. “Why should I trust you?”: Explaining the predictions of any classifier. InProceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 1135-1144, 2016.
[26] Dennis Wei, Sanjeeb Dash, Tian Gao, and Oktay G¨unl¨uk. Generalized linear rule models. In Proceedings of the International Conference on Machine Learning, pages 6687-6696, 2019.
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.