Significance, Relevance and Explainability in the Machine Learning Age: An Econometrics and Financial Data Science Perspective

Forthcoming, European Journal of Finance

14 Pages Posted: 15 Dec 2020

See all articles by Andreas G. F. Hoepner

Andreas G. F. Hoepner

Smurfit Graduate Business School, University College Dublin; European Commission's Platform on Sustainable Finance

David G. McMillan

University of Stirling

Andrew Vivian

Loughborough University

Chardin Wese Simen

University of Liverpool Management School

Date Written: September 30, 2020

Abstract

Although machine learning is frequently associated with neural networks, it also comprises econometric regression approaches and other statistical technique whose accuracy enhances with increasing observation. What constitutes high quality machine learning is yet unclear though. Proponents of deep learning (i.e. neural networks) value computation efficiency over human interpretability and tolerate the “black box” appeal of their algorithms, whereas proponents of explainable artificial intelligence (xai) employ traceable “white box” methods (e.g. regressions) to enhance explainability to human decision makers. We extend Brooks et al.’s (2019) work on significance and relevance as assessment critieria in econometrics and financial data science to contribute to this debate. Specifically, we identify explainability as the Achilles heel of classic machine learning approaches such as neural networks, which are not fully replicable, lack transparency and traceability and therefore do not permit any attempts to establish causal inference. We conclude suggesting routes for future research to advance the design and efficiency of “white box” algorithms.

Keywords: explainability, explainable artificial intelligence (xai), neural networks, relevance, regressions, significance

JEL Classification: C40, C45, C58, C80, Y80

Suggested Citation

Hoepner, Andreas G. F. and McMillan, David G. and Vivian, Andrew and Wese Simen, Chardin, Significance, Relevance and Explainability in the Machine Learning Age: An Econometrics and Financial Data Science Perspective (September 30, 2020). Forthcoming, European Journal of Finance, Available at SSRN: https://ssrn.com/abstract=3706929

Andreas G. F. Hoepner (Contact Author)

Smurfit Graduate Business School, University College Dublin ( email )

Blackrock, Co. Dublin
Ireland

European Commission's Platform on Sustainable Finance ( email )

2 Rue de Spa
Brussels, 1000
Belgium

David G. McMillan

University of Stirling ( email )

Stirling, Scotland FK9 4LA
United Kingdom

Andrew Vivian

Loughborough University ( email )

The Business School
Ashby Road
Loughborough LE11 3TU
Great Britain

Chardin Wese Simen

University of Liverpool Management School ( email )

Management School
University of Liverpool
Liverpool, L69 7ZH
United Kingdom

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
45
Abstract Views
250
PlumX Metrics