Transparency Promotion with Model-Agnostic Linear Competitors

Hassan Rafique, Tong Wang, Qihang Lin, Arshia Singhani
Proceedings of the 37th International Conference on Machine Learning, PMLR 119:7898-7908, 2020.

Abstract

We propose a novel type of hybrid model for multi-class classification, which utilizes competing linear models to collaborate with an existing black-box model, promoting transparency in the decision-making process. Our proposed hybrid model, Model-Agnostic Linear Competitors (MALC), brings together the interpretable power of linear models and the good predictive performance of the state-of-the-art black-box models. We formulate the training of a MALC model as a convex optimization problem, optimizing the predictive accuracy and transparency (defined as the percentage of data captured by the linear models) in the objective function. Experiments show that MALC offers more model flexibility for users to balance transparency and accuracy, in contrast to the currently available choice of either a pure black-box model or a pure interpretable model. The human evaluation also shows that more users are likely to choose MALC for this model flexibility compared with interpretable models and black-box models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v119-rafique20a, title = {Transparency Promotion with Model-Agnostic Linear Competitors}, author = {Rafique, Hassan and Wang, Tong and Lin, Qihang and Singhani, Arshia}, booktitle = {Proceedings of the 37th International Conference on Machine Learning}, pages = {7898--7908}, year = {2020}, editor = {III, Hal Daumé and Singh, Aarti}, volume = {119}, series = {Proceedings of Machine Learning Research}, month = {13--18 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v119/rafique20a/rafique20a.pdf}, url = {https://proceedings.mlr.press/v119/rafique20a.html}, abstract = {We propose a novel type of hybrid model for multi-class classification, which utilizes competing linear models to collaborate with an existing black-box model, promoting transparency in the decision-making process. Our proposed hybrid model, Model-Agnostic Linear Competitors (MALC), brings together the interpretable power of linear models and the good predictive performance of the state-of-the-art black-box models. We formulate the training of a MALC model as a convex optimization problem, optimizing the predictive accuracy and transparency (defined as the percentage of data captured by the linear models) in the objective function. Experiments show that MALC offers more model flexibility for users to balance transparency and accuracy, in contrast to the currently available choice of either a pure black-box model or a pure interpretable model. The human evaluation also shows that more users are likely to choose MALC for this model flexibility compared with interpretable models and black-box models.} }
Endnote
%0 Conference Paper %T Transparency Promotion with Model-Agnostic Linear Competitors %A Hassan Rafique %A Tong Wang %A Qihang Lin %A Arshia Singhani %B Proceedings of the 37th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2020 %E Hal Daumé III %E Aarti Singh %F pmlr-v119-rafique20a %I PMLR %P 7898--7908 %U https://proceedings.mlr.press/v119/rafique20a.html %V 119 %X We propose a novel type of hybrid model for multi-class classification, which utilizes competing linear models to collaborate with an existing black-box model, promoting transparency in the decision-making process. Our proposed hybrid model, Model-Agnostic Linear Competitors (MALC), brings together the interpretable power of linear models and the good predictive performance of the state-of-the-art black-box models. We formulate the training of a MALC model as a convex optimization problem, optimizing the predictive accuracy and transparency (defined as the percentage of data captured by the linear models) in the objective function. Experiments show that MALC offers more model flexibility for users to balance transparency and accuracy, in contrast to the currently available choice of either a pure black-box model or a pure interpretable model. The human evaluation also shows that more users are likely to choose MALC for this model flexibility compared with interpretable models and black-box models.
APA
Rafique, H., Wang, T., Lin, Q. & Singhani, A.. (2020). Transparency Promotion with Model-Agnostic Linear Competitors. Proceedings of the 37th International Conference on Machine Learning, in Proceedings of Machine Learning Research 119:7898-7908 Available from https://proceedings.mlr.press/v119/rafique20a.html.

Related Material