Understanding from Machine Learning Models

British Journal for the Philosophy of Science 73 (1):109-133 (2022)
  Copy   BIBTEX


Simple idealized models seem to provide more understanding than opaque, complex, and hyper-realistic models. However, an increasing number of scientists are going in the opposite direction by utilizing opaque machine learning models to make predictions and draw inferences, suggesting that scientists are opting for models that have less potential for understanding. Are scientists trading understanding for some other epistemic or pragmatic good when they choose a machine learning model? Or are the assumptions behind why minimal models provide understanding misguided? In this paper, using the case of deep neural networks, I argue that it is not the complexity or black box nature of a model that limits how much understanding the model provides. Instead, it is a lack of scientific and empirical evidence supporting the link that connects a model to the target phenomenon that primarily prohibits understanding.

Similar books and articles

Understanding (With) Toy Models.Alexander Reutlinger, Dominik Hangleiter & Stephan Hartmann - 2016 - British Journal for the Philosophy of Science:axx005.
Understanding (with) Toy Models.Alexander Reutlinger, Dominik Hangleiter & Stephan Hartmann - 2018 - British Journal for the Philosophy of Science 69 (4):1069-1099.
Models and Method.D. W. Theobald - 1964 - Philosophy 39 (149):260 - 267.
Understanding does not depend on (causal) explanation.Philippe Verreault-Julien - 2019 - European Journal for Philosophy of Science 9 (2):18.
Explaining Explanations in AI.Brent Mittelstadt - forthcoming - FAT* 2019 Proceedings 1.
How could models possibly provide how-possibly explanations?Philippe Verreault-Julien - 2019 - Studies in History and Philosophy of Science Part A 73:1-12.
Understanding with theoretical models.Petri Ylikoski & N. Emrah Aydinonat - 2014 - Journal of Economic Methodology 21 (1):19-36.


Added to PP

1,169 (#6,423)

6 months
194 (#4,858)

Historical graph of downloads
How can I increase my downloads?

Author's Profile

Emily Sullivan
Eindhoven University of Technology

References found in this work

Depth: An Account of Scientific Explanation.Michael Strevens - 2008 - Cambridge, Mass.: Harvard University Press.
Understanding Why.Alison Hills - 2015 - Noûs 49 (2):661-688.
Inductive risk and values in science.Heather Douglas - 2000 - Philosophy of Science 67 (4):559-579.

View all 30 references / Add more references