The Clinician and Dataset Shift in Artificial Intelligence

This letter outlines how to identify, and potentially mitigate, common sources of “dataset shift” in machine-learning systems. This occurs when the model “training data” differ from the data used by the model to provide diagnostic, prognostic, or treatment advice.

Saved in:
Bibliographic Details
Published in:The New England journal of medicine Vol. 385; no. 3; pp. 283 - 286
Main Authors: Finlayson, Samuel G, Subbaswamy, Adarsh, Singh, Karandeep, Bowers, John, Kupke, Annabel, Zittrain, Jonathan, Kohane, Isaac S, Saria, Suchi
Format: Journal Article
Language:English
Published: United States Massachusetts Medical Society 15.07.2021
Subjects:
Accuracy
Algorithms
and Education
Artificial Intelligence
Calibration
Clinical medicine
Clinical Practice
Clinical trials
COVID-19
Datasets
Datasets as Topic
Demographics
Health IT
Health Policy
Hospitalist
Hospitals
Humans
Influenza
Information Technology
Learning algorithms
Machine Learning
Medical Informatics
Medical Practice
Patients
Primary Care
Quality of Care
Reimbursement
Sepsis
Software
Training
Trends
ISSN:0028-4793, 1533-4406, 1533-4406
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This letter outlines how to identify, and potentially mitigate, common sources of “dataset shift” in machine-learning systems. This occurs when the model “training data” differ from the data used by the model to provide diagnostic, prognostic, or treatment advice.
Bibliography:SourceType-Scholarly Journals-1
ObjectType-Correspondence-1
content type line 14
ObjectType-Article-2
content type line 23
Dr. Finlayson and Mr. Subbaswamy contributed equally to this letter.
ISSN:0028-4793
1533-4406
1533-4406
DOI:10.1056/NEJMc2104626