Year: 2023
Type: preprint
Abstract: Feature importance methods promise to provide a ranking of features according to importance for a given classification task. A wide range of methods exist but their rankings often disagree and they are inherently difficult to evaluate due to a lack of ground truth beyond synthetic datasets. In this work, we put feature importance methods to the test on real-world data in the domain of cardiology, where we try to distinguish three specific pathologies from healthy subjects based on ECG features comparing to features used in cardiologists' decision rules as ground truth. Some methods generally performed well and others performed poorly, while some methods did well on some but not all of the problems considered.
Source: arXiv (Cornell University)
Institutions: Physikalisch-Technische Bundesanstalt, Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute, National Physical Laboratory, Institute of Metrology of Bosnia and Herzegovina, University of Edinburgh, Karlsruhe Institute of Technology, Carl von Ossietzky Universität Oldenburg, University of Surrey
Language: English
Topic: ECG Monitoring and Analysis
Subfield: Cardiology and Cardiovascular Medicine
Field: Medicine
Domain: Health Sciences
Open Access status: green
Awards: G264225365, G434337789
OpenAlex ID: W4362679626