The application of time-to-event analysis in machine learning prognostic models

Unlike conventional classification paradigms, survival analysis confronts the intricacies engendered by partially observed data, often stemming from censoring. Within the realm of clinical inquiries, patient records manifest in distinct categories: those that remain uncensored, thereby divulging pre...

Full description

Saved in:
Bibliographic Details
Published inJournal of translational medicine Vol. 22; no. 1; pp. 146 - 2
Main Authors Peng, Zi-He, Huang, Zhi-Xin, Tian, Juan-Hua, Chong, Tie, Li, Zhao-Lun
Format Journal Article
LanguageEnglish
Published England BioMed Central Ltd 12.02.2024
BioMed Central
BMC
Subjects
Artificial intelligence
Bias
Letter to the Editor
Machine Learning
Medical prognosis
Medical records
Prognosis
Regression analysis
Statistical analysis
Survival analysis
Systematic review
Online AccessGet full text

Cover

More Information
Summary:Unlike conventional classification paradigms, survival analysis confronts the intricacies engendered by partially observed data, often stemming from censoring. Within the realm of clinical inquiries, patient records manifest in distinct categories: those that remain uncensored, thereby divulging precise event timings, and those that exist as right-censored, withholding event timings beyond the study’s temporal scope. Careful consideration of censoring and time-to-event analysis principles is warranted.
Bibliography:content type line 23
SourceType-Scholarly Journals-1
ObjectType-Correspondence-1
ISSN:1479-5876
1479-5876
DOI:10.1186/s12967-024-04909-1