Machine Learning that Matters

TL;DR

This paper critiques current machine learning research for its disconnect from societal and scientific relevance, proposing six Impact Challenges to steer the field towards more meaningful and impactful work.

Contribution

It introduces six Impact Challenges to realign machine learning research with societal and scientific importance, addressing current limitations in data, metrics, and communication.

Findings

Identification of key limitations in current ML research

Proposal of six Impact Challenges to guide future work

Call for increased focus on impactful ML applications

Abstract

Much of current machine learning (ML) research has lost its connection to problems of import to the larger world of science and society. From this perspective, there exist glaring limitations in the data sets we investigate, the metrics we employ for evaluation, and the degree to which results are communicated back to their originating domains. What changes are needed to how we conduct research to increase the impact that ML has? We present six Impact Challenges to explicitly focus the field?s energy and attention, and we discuss existing obstacles that must be addressed. We aim to inspire ongoing discussion and focus on ML that matters.