Abstract
Today the state-of-the-art performance in classification is achieved by the so-called “black boxes”, i.e. decision-making systems whose internal logic is obscure. Such models could revolutionize the health-care system, however their deployment in real-world diagnosis decision support systems is subject to several risks and limitations due to the lack of transparency. The typical classification problem in health-care requires a multi-label approach since the possible labels are not mutually exclusive, e.g. diagnoses. We propose MARLENA, a model-agnostic method which explains multi-label black box decisions. MARLENA explains an individual decision in three steps. First, it generates a synthetic neighborhood around the instance to be explained using a strategy suitable for multi-label decisions. It then learns a decision tree on such neighborhood and finally derives from it a decision rule that explains the black box decision. Our experiments show that MARLENA performs well in terms of mimicking the black box behavior while gaining at the same time a notable amount of interpretability through compact decision rules, i.e. rules with limited length.
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 subscriptionsNotes
- 1.
- 2.
For both neighborhood generation approaches mixed and union, the size of the synthetic neighborhood is 1000, and the size of the core real neighborhood \(X^*\) is \(k = 0.5 |\hat{X}|^{1/2}\).
- 3.
Source code, datasets, and the scripts for reproducing experiments are publicly available at https://github.com/riccotti/ExplainMultilabelClassifiers.
- 4.
- 5.
We replace the missing values with the mean for continuous variables and with the mode for categorical ones. We remove the features with more than 40% of missing values.
- 6.
Implementations are those of scikit-learn library.
- 7.
Details available at https://github.com/riccotti/ExplainMultilabelClassifiers.
- 8.
The performance reported consider only instances for which an explanation is returned. Indeed, for some instances of the medical dataset using the RF black box an explanation is not returned. We leave the investigation of this specific case fur future studies.
References
Abe, S.: Fuzzy support vector machines for multilabel classification. PR 48(6), 2110 (2015)
Bai, T., et al.: Interpretable representation learning for healthcare via capturing disease progression through time. In: KDD, pp. 43–51. ACM (2018)
Blockeel, H., Schietgat, L., Struyf, J., Clare, A., Dzeroski, S.: Hierarchical multilabel classification trees for gene function prediction. In: MLSB, pp. 9–14 (2006)
Che, Z., Kale, D., Li, W., Bahadori, M.T., Liu, Y.: Deep computational phenotyping. In: KDD, pp. 507–516. ACM (2015)
Che, Z., Purushotham, S., Khemani, R., Liu, Y.: Interpretable deep models for ICU outcome prediction. In: AMIA Annual Symposium Proceedings, vol. 2016, p. 371. American Medical Informatics Association (2017)
Choi, E., et al.: Doctor AI: predicting clinical events via recurrent neural networks. In: Machine Learning for Healthcare Conference, pp. 301–318 (2016)
Chui, M.: Artificial intelligence the next digital frontier? McKinsey and CGI, p. 47 (2017)
Elisseeff, A., Weston, J.: A kernel method for multi-labelled classification. In: Advances in Neural Information Processing Systems, pp. 681–687 (2002)
Feng, Y., et al.: Patient outcome prediction via convolutional neural networks based on multi-granularity medical concept embedding. In: BIBM, pp. 770–777. IEEE (2017)
Guidotti, R., Monreale, A., Ruggieri, S., Pedreschi, D., Turini, F., Giannotti, F.: Local rule-based explanations of black box decision systems. CoRR (2018). arXiv:abs/1805.10820
Guidotti, R., Monreale, A., Ruggieri, S., Turini, F., Pedreschi, D., Giannotti, F.: A survey of methods for explaining black box models. ACM CSUR 51(5), 93:1–93:42 (2018)
Guidotti, R., Soldani, J., Neri, D., Brogi, A., Pedreschi, D.: Helping your Docker images to spread based on explainable models. In: ECML-PKDD. Springer, Berlin (2018)
Lasko, T.A., et al.: Computational phenotype discovery using unsupervised feature learning over noisy, sparse, and irregular clinical data. PloS One 8(6), e66341 (2013)
LeCun, Y., Bengio, Y., Hinton, G.: Deep learning. Nature 521(7553), 436 (2015)
Malgieri, G., Comandé, G.: Why a right to legibility of automated decision-making exists in the general data protection regulation. Int. Data Priv. Law 7(4), 243–265 (2017)
Miotto, R., et al.: Deep patient: an unsupervised representation to predict the future of patients from the electronic health records. Sci. Rep. 6, 26094 (2016)
Pestian, J.P., et al.: A shared task involving multi-label classification of clinical free text. In: BioNLP, pp. 97–104. Association for Computational Linguistics (2007)
Rajkomar, A., et al.: Scalable and accurate deep learning with EHR. DM 1(1), 18 (2018)
Ribeiro, M.T., Singh, S., Guestrin, C.: Why should i trust you?: explaining the predictions of any classifier. In: KDD, pp. 1135–1144. ACM (2016)
Sapozhnikova, E.P.: Art-based neural networks for multi-label classification. In: International Symposium on Intelligent Data Analysis, pp. 167–177. Springer, Berlin (2009)
Shickel, B., et al.: Deep EHR: a survey of recent advances in deep learning techniques for EHR analysis. J. Biomed. Health Inform. 22(5), 1589–1604 (2018)
Tan, P.-N. et al.: Introduction to data mining. Pearson Education India (2007)
Tsoumakas, G., Katakis, I.: Multi-label classification: an overview. Int. J. Data Warehous. Min. (IJDWM) 3(3), 1–13 (2007)
Vens, C., Struyf, J., Schietgat, L., Džeroski, S., Blockeel, H.: Decision trees for hierarchical multi-label classification. Mach. Learn. 73(2), 185 (2008)
Wachter, S., et al.: Why a right to explanation of automated decision-making does not exist in the general data protection regulation. Int. Data Priv. Law 7(2), 76–99 (2017)
Yadav, P., Steinbach, M., Kumar, V., Simon, G.: Mining electronic health records (EHRS): a survey. ACM Comput. Surv. (CSUR) 50(6), 85 (2018)
Acknowledgements
This work is partially supported by the European H2020 Program under the funding scheme “INFRAIA-1-2014-2015: Research Infrastructures” g.a. 654024 “SoBigData”, http://www.sobigdata.eu.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Panigutti, C., Guidotti, R., Monreale, A., Pedreschi, D. (2020). Explaining Multi-label Black-Box Classifiers for Health Applications. In: Shaban-Nejad, A., Michalowski, M. (eds) Precision Health and Medicine. W3PHAI 2019. Studies in Computational Intelligence, vol 843. Springer, Cham. https://doi.org/10.1007/978-3-030-24409-5_9
Download citation
DOI: https://doi.org/10.1007/978-3-030-24409-5_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-24408-8
Online ISBN: 978-3-030-24409-5
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)