PhilSci Archive

Machine Learning models as Mathematics: interpreting explainable AI in non-causal terms

Buijsman, Stefan (2024) Machine Learning models as Mathematics: interpreting explainable AI in non-causal terms. [Preprint]

[img]
Preview
Text
Machine_Learning_models_as_Mathematics.pdf

Download (131kB) | Preview

Abstract

We would like to have a wide range of explanations for the behaviour of machine learning systems. However, how should we understand these explanations? Typically, attempts to clarify what an explanations for questions such as ’why am I getting this output for these inputs?’ have been approached from the philosophy of science, through an analogy with scientific (and often causal) explanations. I show that ML systems are best thought of as non-causal, specifically mathematical objects. We should therefore interpret these explanations differently, through analogy with mathematical explanations. I show that this still allows us to use much of the same theoretical apparatus, and argue that the asymmetry of many of the standard ML explanations can be accounted for in virtue of the link these systems have with concrete implementations.


Export/Citation: EndNote | BibTeX | Dublin Core | ASCII/Text Citation (Chicago) | HTML Citation | OpenURL
Social Networking:
Share |

Item Type: Preprint
Creators:
CreatorsEmailORCID
Buijsman, Stefans.n.r.buijsman@tudelft.nl0000-0002-0004-0681
Keywords: Explainable AI; Machine learning; Interpretability; Non-causal explanation; Mathematical explanation
Subjects: Specific Sciences > Mathematics > Explanation
Specific Sciences > Artificial Intelligence
General Issues > Explanation
Specific Sciences > Artificial Intelligence > Machine Learning
Depositing User: Dr. Stefan Buijsman
Date Deposited: 16 Mar 2024 01:47
Last Modified: 16 Mar 2024 01:47
Item ID: 23201
Subjects: Specific Sciences > Mathematics > Explanation
Specific Sciences > Artificial Intelligence
General Issues > Explanation
Specific Sciences > Artificial Intelligence > Machine Learning
Date: 2024
URI: https://philsci-archive.pitt.edu/id/eprint/23201

Monthly Views for the past 3 years

Monthly Downloads for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item