Comparison of the Cox proportional hazards model and Random Survival Forest algorithm for predicting patient-specific survival probabilities in clinical trial data

Ricarda Graf, Susan Todd, M. Fazil Baksh · 2025

Read Comparison of the Cox proportional hazards model and Random Survival Forest algorithm for predicting patient-specific survival probabilities in clinical trial data by Ricarda Graf, Susan Todd, M. Fazil Baksh (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-03119.txt.

Description

Ricarda Graf, Susan Todd, and Fazil Baksh compare two ways to predict a patient's chance of survival over time from trial data, in a paper from 2025. The classic tool is the Cox model, a workhorse of medical statistics. The challenger is a machine learning method called random survival forest, praised for strong accuracy.

Their comparison was built to be neutral, with simulated scenarios drawn from two real trial data sets. Performance was judged by the reporting standards used for prediction models, not by a single score. The authors show that leaning on one popular accuracy index can mislead the choice between the methods. They map which settings favor which tool, giving trial analysts a fairer basis for choosing.

Catalog record

CREATORRicarda Graf, Susan Todd, M. Fazil Baksh
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE51 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-03119

Browse the full library · About the Tower of Babel