Abstract
Evaluating the explanations given by post-hoc XAI approaches on tabular data is a challenging prospect, since the subjective judgement of explanations of tabular relations is non trivial in contrast to e.g. the judgement of image heatmap explanations. In order to quantify XAI performance on categorical tabular data, where feature relationships can often be described by Boolean functions, we propose an evaluation setting through generation of synthetic datasets. To create gold standard explanations, we present a definition of feature relevance in Boolean functions. In the proposed setting we evaluate eight state-of-the-art XAI approaches and gain novel insights into XAI performance on categorical tabular data. We find that the investigated approaches often fail to faithfully explain even basic relationships within categorical data.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Ancona, M., Ceolini, E., Öztireli, C., Gross, M.: A unified view of gradient-based attribution methods for deep neural networks. In: NIPS 2017 - Workshop on Interpreting, Explaining and Visualizing Deep Learning. ETH Zurich (2017)
Ancona, M., Ceolini, E., Öztireli, C., Gross, M.: Gradient-based attribution methods. In: Samek, W., Montavon, G., Vedaldi, A., Hansen, L.K., Müller, K.-R. (eds.) Explainable AI: Interpreting, Explaining and Visualizing Deep Learning. LNCS (LNAI), vol. 11700, pp. 169–191. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-28954-6_9
Bach, S., Binder, A., Montavon, G., Klauschen, F., Müller, K.R., Samek, W.: On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation. PloS one 10(7), e0130140 (2015)
Castro, J., Gómez, D., Tejada, J.: Polynomial calculation of the shapley value based on sampling. Comput. Oper. Res. 36(5), 1726–1730 (2009)
Kindermans, P.J., et al.: Learning how to explain neural networks: Patternnet and patternattribution. In: International Conference on Learning Representations (2018)
Lundberg, S.M., Lee, S.I.: A unified approach to interpreting model predictions. In: Advances in Neural Information Processing Systems, pp. 4765–4774 (2017)
O’Donnell, R.: Analysis of Boolean Functions. Cambridge University Press, Cambridge (2014)
Preiss, B.: Data Structures and Algorithms with Object-Oriented Design Patterns in Java (1999)
Ribeiro, M.T., Singh, S., Guestrin, C.: Why should i trust you?: explaining the predictions of any classifier. In: 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data mining, pp. 1135–1144. ACM (2016)
Ring, M., Schlr, D., Landes, D., Hotho, A.: Flow-based network traffic generation using generative adversarial networks. Comput. Secur. 82, 156–172 (2019)
Rudin, C.: Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell. 1(5), 206 (2019)
Shrikumar, A., Greenside, P., Kundaje, A.: Learning important features through propagating activation differences. In: 34th International Conference on Machine Learning, vol. 70, pp. 3145–3153. JMLR. org (2017)
Shrikumar, A., Greenside, P., Shcherbina, A., Kundaje, A.: Not just a black box: Learning important features through propagating activation differences. arXiv preprint arXiv:1605.01713 (2016)
Simonyan, K., Vedaldi, A., Zisserman, A.: Deep inside convolutional networks: visualising image classification models and saliency maps. In: Bengio, Y., LeCun, Y. (eds.) ICLR (Workshop Poster) (2014)
Sundararajan, M., Taly, A., Yan, Q.: Axiomatic attribution for deep networks. In: 34th International Conference on Machine Learning, vol. 70, pp. 3319–3328. JMLR. org (2017)
Acknowledgement
The authors acknowledge the financial support by the Federal Ministry of Education and Research of Germany as part of the DeepScan project (01IS18045A).
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Tritscher, J., Ring, M., Schlr, D., Hettinger, L., Hotho, A. (2020). Evaluation of Post-hoc XAI Approaches Through Synthetic Tabular Data. In: Helic, D., Leitner, G., Stettinger, M., Felfernig, A., Raś, Z.W. (eds) Foundations of Intelligent Systems. ISMIS 2020. Lecture Notes in Computer Science(), vol 12117. Springer, Cham. https://doi.org/10.1007/978-3-030-59491-6_40
Download citation
DOI: https://doi.org/10.1007/978-3-030-59491-6_40
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-59490-9
Online ISBN: 978-3-030-59491-6
eBook Packages: Computer ScienceComputer Science (R0)