Reliability and Interpretability in Science and Deep Learning
Open Access
- 25 June 2024
- journal article
- research article
- Published by Springer Nature in Minds and Machines
- Vol. 34 (3) , 1-31
Abstract
In recent years, the question of the reliability of Machine Learning (ML) methods has acquired significant importance, and the analysis of the associated uncertainties has motivated a growing amount of research. However, most of these studies have applied standard error analysis to ML models—and in particular Deep Neural Network (DNN) models—which represent a rather significant departure from standard scientific modelling. It is therefore necessary to integrate the standard error analysis with a deeper epistemological analysis of the possible differences between DNN models and standard scientific modelling and the possible implications of these differences in the assessment of reliability. This article offers several contributions. First, it emphasises the ubiquitous role of model assumptions (both in ML and traditional science) against the illusion of theory-free science. Secondly, model assumptions are analysed from the point of view of their (epistemic) complexity, which is shown to be language-independent. It is argued that the high epistemic complexity of DNN models hinders the estimate of their reliability and also their prospect of long term progress. Some potential ways forward are suggested. Thirdly, this article identifies the close relation between a model’s epistemic complexity and its interpretability, as introduced in the context of responsible AI. This clarifies in which sense—and to what extent—the lack of understanding of a model (black-box problem) impacts its interpretability in a way that is independent of individual skills. It also clarifies how interpretability is a precondition for a plausible assessment of the reliability of any model, which cannot be based on statistical analysis alone. This article focuses on the comparison between traditional scientific models and DNN models. However, Random Forest (RF) and Logistic Regression (LR) models are also briefly considered.Keywords
All Related Versions
This publication has 81 references indexed in Scilit:
- The Explanation Game: A Formal Framework for Interpretable Machine Learning2021
- A review on weight initialization strategies for neural networksArtificial Intelligence Review, 2021
- An Introduction to Kolmogorov Complexity and Its Applications2019
- Fraud Analytics Using Descriptive, Predictive, and Social Network Techniques2015
- What’s Right with a Syntactic Approach to Theories and Models?Erkenntnis, 2014
- Falsificationism and Statistical Learning Theory: Comparing the Popper and Vapnik-Chervonenkis DimensionsJournal for General Philosophy of Science, 2009
- The Elements of Statistical Learning2009
- Patterns of Change2008
- On empirically equivalent systems of the worldErkenntnis, 1975
- On simple theories of a complex worldSynthese, 1963