# A Naive Bayes machine learning approach to risk prediction using censored, time-to-event data

Source: [A Naive Bayes machine learning approach to risk prediction using censored, time-to-event data](<https://www.sota2.com/research/paper/68fa7d2e1e209131cea91d93>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T06:16:39.216Z

Authors: Julian Wolfson, Sunayan Bandyopadhyay, Mohamed Elidrisi, Gabriela Vazquez-Benitez, Donald Musgrove, Gediminas Adomavicius, Paul Johnson, Patrick O'Connor

Published: 2014-04-08T13:34:29.000Z

## Abstract

Predicting an individual's risk of experiencing a future clinical outcome is a statistical task with important consequences for both practicing clinicians and public health experts. Modern observational databases such as electronic health records (EHRs) provide an alternative to the longitudinal cohort studies traditionally used to construct risk models, bringing with them both opportunities and challenges. Large sample sizes and detailed covariate histories enable the use of sophisticated machine learning techniques to uncover complex associations and interactions, but observational databases are often \`\`messy,'' with high levels of missing data and incomplete patient follow-up. In this paper, we propose an adaptation of the well-known Naive Bayes (NB) machine learning approach for classification to time-to-event outcomes subject to censoring. We compare the predictive performance of our method to the Cox proportional hazards model which is commonly used for risk prediction in healthcare populations, and illustrate its application to prediction of cardiovascular risk using an EHR dataset from a large Midwest integrated healthcare system.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Clinical Outcome Prediction on MIMIC-IV](<https://www.sota2.com/research/sota/clinical-outcome-prediction-on-mimic-iv.md>): Clinical Outcome Prediction; MIMIC IV; Mortality (dCorr): 32.09.

- [ICU Transfer Prediction on MIMIC-III](<https://www.sota2.com/research/sota/icu-transfer-prediction-on-mimic-iii.md>): ICU transfer prediction (ICU); MIMIC-III; MI (x10^3): 27.53.

- [30-day Mortality Prediction (30M) on MIMIC-IV](<https://www.sota2.com/research/sota/30-day-mortality-prediction-30m-on-mimic-iv.md>): 30-day mortality prediction (30M); MIMIC IV; Mutual Information (MI): 0.05774.

- [In-hospital mortality prediction on MIMIC-IV (using MI)](<https://www.sota2.com/research/sota/in-hospital-mortality-prediction-on-mimic-iv-using-mi.md>): In-hospital mortality prediction (MORT); MIMIC IV; MI (x10^3): 56.68.

- [30-day mortality prediction (30M) on MIMIC-III](<https://www.sota2.com/research/sota/30-day-mortality-prediction-30m-on-mimic-iii.md>): 30-day mortality prediction (30M); MIMIC-III; Mutual Information (MI) (x10^3): 55.43.

- [Clinical Outcome Prediction on MIMIC-III](<https://www.sota2.com/research/sota/clinical-outcome-prediction-on-mimic-iii.md>): Clinical Outcome Prediction; MIMIC-III; Mortality (dCorr): 24.89.

- [ICU Transfer Prediction on MIMIC-IV](<https://www.sota2.com/research/sota/icu-transfer-prediction-on-mimic-iv.md>): ICU transfer prediction (ICU); MIMIC IV; MI (x10^2): 13.37.

- [In-hospital mortality prediction (MORT) on MIMIC-III](<https://www.sota2.com/research/sota/in-hospital-mortality-prediction-mort-on-mimic-iii.md>): In-hospital mortality prediction (MORT); MIMIC-III; Mutual Information (MI): 0.05775.

- [Length of Stay Prediction (LOS) on MIMIC-IV (Mutual Information)](<https://www.sota2.com/research/sota/length-of-stay-prediction-los-on-mimic-iv-mutual-information.md>): Length of Stay Prediction (LOS); MIMIC IV; Mutual Information (MI): 0.079.

- [Length of stay prediction (LOS) on MIMIC-III](<https://www.sota2.com/research/sota/length-of-stay-prediction-los-on-mimic-iii.md>): Length of Stay Prediction (LOS); MIMIC-III; Mutual Information (MI): 0.0727.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1404.2124>)

[PDF](<https://arxiv.org/pdf/1404.2124.pdf>)