Bootstrap Methods for Developing Predictive Models

2004/05/01 by Peter C. Austin, Peter C Austin, Jack V Tu +1 · 30 citations
Environmental Science · Mathematics · #Hydrology and Drought Analysis #Statistical Methods and Inference #Advanced Statistical Methods and Models

paper · doi:10.1198/0003130043277

Abstract

Researchers frequently use automated model selection methods such as backwards elimination to identify variables that are independent predictors of an outcome under consideration. We propose using bootstrap resampling in conjunction with automated variable selection methods to develop parsimonious prediction models. Using data on patients admitted to hospital with a heart attack, we demonstrate that selecting those variables that were identified as independent predictors of mortality in at least 60%% of the bootstrap samples resulted in a parsimonious model with excellent predictive ability.

Cited by

• Survival prediction for stage I-IIIA non-small cell lung cancer using deep learning
• Nonparametric Causal Feature Selection for Spatiotemporal Risk Mapping of Malaria Incidence in Madagascar
• Super El Niño events drive climate regime shifts with enhanced risks under global warming
• Modification of the Thrombolysis in Myocardial Infarction risk score for patients presenting with chest pain to the emergency department
• Interoceptive accuracy and cognitive reappraisal modulate affective and physiological reactivity to social stress
• Determining relative importance of variables in developing and validating predictive models. [europepmc]
• Projected range contractions of montane biodiversity under global warming. [europepmc]
• The search for stable prognostic models in multiple imputed data sets. [europepmc]
• Does hyperthermic intraoperative chemotherapy lead to improved outcomes in patients with ovarian cancer? A single center cohort study in 111 consecutive patients. [europepmc]
• A risk score for identifying methicillin-resistant Staphylococcus aureus in patients presenting to the hospital with pneumonia. [europepmc]
• Quality of embryos transferred and progesterone levels are the most important predictors of live birth after fresh embryo transfer: a retrospective cohort study. [europepmc]
• A public-private partnership develops and externally validates a 30-day hospital readmission risk prediction model. [europepmc]
• The diagnostic value of the combination of patient characteristics, history, and clinical shoulder tests for the diagnosis of rotator cuff tear. [europepmc]
• Comprehensive analysis and selection of anthrax vaccine adsorbed immune correlates of protection in rhesus macaques. [europepmc]
• Prediction of treatment response at 5-year follow-up in a randomized clinical trial of behaviorally based couple therapies. [europepmc]
• Reporting and handling of missing data in predictive research for prevalent undiagnosed type 2 diabetes mellitus: a systematic review. [europepmc]
• A multiplex urinary immunoassay for bladder cancer detection: analysis of a Japanese cohort. [europepmc]
• Development and validation of a clinical prediction model for patient-reported pain and function after primary total knee replacement surgery. [europepmc]
• Factors influencing temporal patterns in crime in a large American city: A predictive analytics perspective. [europepmc]
• Development and Validation of a Risk Prediction Model for In-Hospital Mortality Among Patients With Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis-ABCD-10. [europepmc]
• Combining Individual Phenotypes of Feed Intake With Genomic Data to Improve Feed Efficiency in Sea Bass. [europepmc]
• Medication-related harm in older adults following hospital discharge: development and validation of a prediction tool. [europepmc]
• Factors That Influence Conversion to Resectability and Survival After Resection of Metastases in RAS WT Metastatic Colorectal Cancer (mCRC): Analysis of FIRE-3- AIOKRK0306. [europepmc]
• Development and Validation of a Penicillin Allergy Clinical Decision Rule. [europepmc]
• AutoScore: A Machine Learning-Based Automatic Clinical Score Generator and Its Application to Mortality Prediction Using Electronic Health Records. [europepmc]
• Neural oscillations underlying selective attention follow sexually divergent developmental trajectories during adolescence. [europepmc]
• Neutrophils and Lymphocytes Can Help Distinguish Asymptomatic COVID-19 From Moderate COVID-19. [europepmc]
• Development and External Validation of a Mortality Prediction Model for Community-Dwelling Older Adults With Dementia. [europepmc]
• The development of sensorimotor cortical oscillations is mediated by pubertal testosterone. [europepmc]
• Left Atrial Mechanical Dysfunction and the Risk for Ischemic Stroke in People Without Prevalent Atrial Fibrillation or Stroke : A Prospective Cohort Study. [europepmc]

Related

• Improving Model Robustness with Bootstrapping
• A Machine Learning Alternative to P-values
• Comparing methods addressing multi-collinearity when developing prediction models
• Variable selection strategies and its importance in clinical prediction modelling
• Stable variable selection for right censored data: comparison of methods