Internal validation strategy for high dimensional prognosis model: A simulation study and application to transcriptomic in head and neck tumors

Abstract excerpt

Background: Predictive models using high-dimensional data, such as genomics and transcriptomics, are increasingly used in oncology for time-to-event endpoints. Internal validation of these models is crucial to mitigate optimism bias prior to external validation. Common strategies include train-test, bootstrap, and (nested) cross-validation. However, no benchmark exists for these methods in high-dimensional...

Topics

Share this publication in a Topic to start or enrich a Post.

• Share to Ferroptosis and cancer prognosis
• Share to Gene expression and cancer classification
• Share to Head and Neck Cancer Studies