Ratti, Emanuele and Termine, Alberto and Facchini, Alessandro (2024) Machine Learning and Theory-Ladenness: A Phenomenological Account. [Preprint]

Text ML and theory ladenness draft July.pdf Download (479kB)

Abstract

In recent years, the dissemination of machine learning (ML) methodologies in scientific research has prompted discussions on theory-ladenness. More specifically, the issue of theory-ladenness has re-emerged as questions about whether and how ML models (MLMs) and ML modelling strategies are impacted by the domain theory of the scientific field in which ML is used and implemented (e.g., physics, chemistry, biology, etc). On the one hand, some have argued that there is no difference between ‘traditional’ (pre-ML) and ML-assisted science. In both cases, theory plays an essential and unavoidable role in the analysis of phenomena and the construction and use of models. Others have argued instead that ML methodologies and models are theory-independent and, in some cases, even theory-free. In this article, we argue that both positions are overly simplistic and do not advance our understanding of the interplay between ML methods and domain theories. Specifically, we provide an analysis of theory-ladenness in ML-assisted science. We do so by constructing an account of MLMs based on a comparison with phenomenological models (PMs), and we show that seeing MLMs through the lens of the debate on PMs, can shed light on the subtle roles that domain-theory plays in the various steps of the construction and use of MLMs. Our analysis reveals that, while the construction of MLMs can be relatively independent of domain-theory, the practical implementation and interpretation of these models within a given specific domain still relies on fundamental theoretical assumptions and background knowledge. Based on our analysis, we introduce new categories of theory-ladenness - 'theory indifference,' 'theory-aid,' and 'theory-infection’ - to capture the varying degrees of influence of domain-theory on MLMs. This analysis of theory-ladenness has far-reaching consequences for understanding the role of ML practices in contemporary science, and the relation between ML specialists and scientists of a given field

---

Export/Citation:

EndNote | BibTeX | Dublin Core | ASCII/Text Citation (Chicago) | HTML Citation | OpenURL

Social Networking:

Share |

---

Item Type:	Preprint
Creators:	Creators Email ORCID Ratti, Emanuele mnl.ratti@gmail.com 0000-0003-1409-8240 Termine, Alberto alberto.termine@unimi.it 0000-0001-5993-0948 Facchini, Alessandro alessandro.facchini@idsia.ch 0000-0001-7507-116X
Keywords:	machine learning theory-ladenness phenomenological models scientific modelling
Subjects:	Specific Sciences > Artificial Intelligence > Machine Learning General Issues > Models and Idealization General Issues > Technology General Issues > Theory Change General Issues > Theory/Observation
Depositing User:	Dr Emanuele Ratti
Date Deposited:	05 Sep 2024 12:51
Last Modified:	05 Sep 2024 12:51
Item ID:	23876
Subjects:	Specific Sciences > Artificial Intelligence > Machine Learning General Issues > Models and Idealization General Issues > Technology General Issues > Theory Change General Issues > Theory/Observation
Date:	2024
URI:	https://philsci-archive.pitt.edu/id/eprint/23876

Monthly Views for the past 3 years

Monthly Downloads for the past 3 years

Plum Analytics

Actions (login required)

View Item